This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2011-02-23 02:38:13 +08:00
|
|
|
! Copyright (C) 2008-2011 Quantum ESPRESSO group
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
! This file is distributed under the terms of the
|
|
|
|
! GNU General Public License. See the file `License'
|
|
|
|
! in the root directory of the present distribution,
|
|
|
|
! or http://www.gnu.org/copyleft/gpl.txt .
|
|
|
|
!
|
|
|
|
!=----------------------------------------------------------------------------=!
|
|
|
|
MODULE read_upf_v2_module
|
|
|
|
!=----------------------------------------------------------------------------=!
|
|
|
|
! this module handles the reading of pseudopotential data
|
|
|
|
|
|
|
|
! ... declare modules
|
|
|
|
USE kinds, ONLY: DP
|
|
|
|
USE pseudo_types, ONLY: pseudo_upf
|
|
|
|
USE radial_grids, ONLY: radial_grid_type
|
2008-04-24 23:26:26 +08:00
|
|
|
USE parser, ONLY : version_compare
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
USE iotk_module
|
|
|
|
!
|
|
|
|
PRIVATE
|
|
|
|
PUBLIC :: read_upf_v2
|
|
|
|
CONTAINS
|
|
|
|
|
|
|
|
!------------------------------------------------+
|
|
|
|
SUBROUTINE read_upf_v2(u, upf, grid, ierr) !
|
|
|
|
!---------------------------------------------+
|
|
|
|
! Read pseudopotential in UPF format version 2, uses iotk
|
|
|
|
!
|
|
|
|
USE pseudo_types, ONLY: nullify_pseudo_upf, deallocate_pseudo_upf
|
|
|
|
USE radial_grids, ONLY: radial_grid_type, nullify_radial_grid
|
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
TYPE(radial_grid_type),OPTIONAL,INTENT(INOUT),TARGET :: grid
|
|
|
|
!
|
|
|
|
INTEGER,OPTIONAL,INTENT(OUT):: ierr ! /= 0 if something went wrong
|
|
|
|
CHARACTER(len=iotk_namlenx) :: root
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
2008-04-09 22:51:58 +08:00
|
|
|
INTEGER :: ierr_
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
LOGICAL :: found
|
|
|
|
LOGICAL,EXTERNAL :: matches
|
2009-04-03 00:11:13 +08:00
|
|
|
CHARACTER(len=6),PARAMETER :: max_version = '2.0.1'
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-27 16:58:55 +08:00
|
|
|
! Prepare the type . Should be done where upf is instantiated
|
|
|
|
! CALL deallocate_pseudo_upf(upf)
|
|
|
|
! CALL nullify_pseudo_upf(upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-27 16:58:55 +08:00
|
|
|
! IF(present(grid)) call nullify_radial_grid(grid)
|
|
|
|
! nullify(upf%grid)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Initialize the file
|
|
|
|
CALL iotk_open_read(u, attr=attr, root=root, ierr=ierr_)
|
|
|
|
!
|
2008-04-11 20:32:27 +08:00
|
|
|
IF((abs(ierr_)>0) .or. .not. matches('UPF',root) ) THEN
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2010-03-27 18:46:09 +08:00
|
|
|
CALL iotk_close_read(u,ierr=ierr_)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF(.not. present(ierr)) &
|
2009-04-03 00:11:13 +08:00
|
|
|
CALL errore('read_upf_v2','Cannot open UPF file.',1)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ierr = 1
|
|
|
|
RETURN
|
|
|
|
ENDIF
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'version', upf%nv)
|
|
|
|
IF (version_compare(upf%nv, max_version) == 'newer') &
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL errore('read_upf_v2',&
|
2008-04-09 22:51:58 +08:00
|
|
|
'Unknown UPF format version: '//TRIM(upf%nv),1)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Skip human-readable header
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_begin(u,'PP_INFO',found=found)
|
2008-04-11 20:32:27 +08:00
|
|
|
if(found) CALL iotk_scan_end(u,'PP_INFO')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read machine-readable header
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_header(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF(upf%tpawp .and. .not. present(grid)) &
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL errore('read_upf_v2', 'PAW requires a radial_grid_type.', 1)
|
2009-04-03 00:11:13 +08:00
|
|
|
!
|
2011-02-23 02:38:13 +08:00
|
|
|
! CHECK for bug in version 2.0.0 of UPF file
|
2009-05-05 15:11:40 +08:00
|
|
|
IF ( version_compare(upf%nv, '2.0.1') == 'older' .and. upf%tvanp .and. &
|
|
|
|
.not. upf%tpawp ) CALL errore('read_upf_v2',&
|
2011-02-23 02:38:13 +08:00
|
|
|
'Ultrasoft pseudopotentials in UPF format v.2.0.0 are &
|
|
|
|
& affected by a bug compromising their quality. Please &
|
|
|
|
& regenerate pseudopotential file for '//TRIM(upf%psd), 1)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read radial grid mesh
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_mesh(u, upf, grid)
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read non-linear core correction charge
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ALLOCATE( upf%rho_atc(upf%mesh) )
|
|
|
|
IF(upf%nlcc) THEN
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_dat(u, 'PP_NLCC', upf%rho_atc)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ELSE
|
|
|
|
! A null core charge simplifies several functions, mostly in PAW
|
|
|
|
upf%rho_atc(1:upf%mesh) = 0._dp
|
|
|
|
ENDIF
|
2010-07-06 20:04:34 +08:00
|
|
|
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read local potential
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF(.not. upf%tcoulombp) THEN
|
|
|
|
ALLOCATE( upf%vloc(upf%mesh) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_LOCAL', upf%vloc)
|
|
|
|
ENDIF
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read nonlocal components: projectors, augmentation, hamiltonian elements
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_nonlocal(u, upf)
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read initial pseudo wavefunctions
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
! (usually only wfcs with occupancy > 0)
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_pswfc(u, upf)
|
2008-04-16 18:12:11 +08:00
|
|
|
! Read all-electron and pseudo wavefunctions
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_full_wfc(u, upf)
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read valence atomic density (used for initial density)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ALLOCATE( upf%rho_at(upf%mesh) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_RHOATOM', upf%rho_at)
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read additional info for full-relativistic calculation
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_spin_orb(u, upf)
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read additional data for PAW (All-electron charge, wavefunctions, vloc..)
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_paw(u, upf)
|
2010-07-06 20:04:34 +08:00
|
|
|
! Read data for gipaw reconstruction
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL read_upf_gipaw(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Close the file (not the unit!)
|
|
|
|
CALL iotk_close_read(u)
|
2010-03-27 18:46:09 +08:00
|
|
|
!
|
|
|
|
IF( present(ierr) ) ierr=0
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
|
|
|
|
RETURN
|
|
|
|
|
|
|
|
CONTAINS
|
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_header(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
2011-05-26 19:53:24 +08:00
|
|
|
CHARACTER(len=256) :: dft_buffer ! needed to allow the string defining the
|
|
|
|
! DFT flavor to be longer than upf%dft
|
|
|
|
! (currntly 25) without getting iotk upset.
|
|
|
|
! An error message is issued if trimmed
|
|
|
|
! dft_buffer exceeds upf%dft size.
|
|
|
|
INTEGER :: len_buffer
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
INTEGER :: nw
|
|
|
|
!
|
|
|
|
! Read HEADER section with some initialization data
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_empty(u, 'PP_HEADER', attr=attr)
|
2010-12-08 00:17:49 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'generated', upf%generated, default=' ')
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'author', upf%author, default='anonymous')
|
2010-12-08 00:17:49 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'date', upf%date, default=' ')
|
|
|
|
CALL iotk_scan_attr(attr, 'comment', upf%comment, default=' ')
|
2008-04-09 22:51:58 +08:00
|
|
|
!
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'element', upf%psd)
|
|
|
|
CALL iotk_scan_attr(attr, 'pseudo_type', upf%typ)
|
|
|
|
CALL iotk_scan_attr(attr, 'relativistic', upf%rel)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'is_ultrasoft', upf%tvanp)
|
|
|
|
CALL iotk_scan_attr(attr, 'is_paw', upf%tpawp)
|
|
|
|
CALL iotk_scan_attr(attr, 'is_coulomb', upf%tcoulombp, default=.false.)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'has_so', upf%has_so, default=.false.)
|
2008-04-16 18:12:11 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'has_wfc', upf%has_wfc, default=upf%tpawp)
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'has_gipaw', upf%has_gipaw, default=.false.)
|
2010-07-06 20:04:34 +08:00
|
|
|
!EMINE
|
|
|
|
CALL iotk_scan_attr(attr, 'paw_as_gipaw', upf%paw_as_gipaw, default=.false.)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'core_correction',upf%nlcc)
|
2011-05-26 19:53:24 +08:00
|
|
|
! CALL iotk_scan_attr(attr, 'functional', upf%dft)
|
|
|
|
CALL iotk_scan_attr(attr, 'functional', dft_buffer)
|
|
|
|
len_buffer=len_trim(dft_buffer)
|
|
|
|
if (len_buffer > len(upf%dft)) &
|
|
|
|
call errore('read_upf_v2','String defining DFT is too long',len_buffer)
|
|
|
|
upf%dft=TRIM(dft_buffer)
|
|
|
|
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'z_valence', upf%zp)
|
|
|
|
CALL iotk_scan_attr(attr, 'total_psenergy', upf%etotps, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'wfc_cutoff', upf%ecutwfc, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'rho_cutoff', upf%ecutrho, default=0._dp)
|
2014-03-15 22:46:02 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'l_max', upf%lmax, default=0)
|
2008-04-11 20:32:27 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'l_max_rho', upf%lmax_rho, default=2*upf%lmax)
|
|
|
|
CALL iotk_scan_attr(attr, 'l_local', upf%lloc, default=0)
|
|
|
|
CALL iotk_scan_attr(attr, 'mesh_size', upf%mesh)
|
|
|
|
CALL iotk_scan_attr(attr, 'number_of_wfc', upf%nwfc)
|
|
|
|
CALL iotk_scan_attr(attr, 'number_of_proj', upf%nbeta)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2008-04-09 22:51:58 +08:00
|
|
|
!CALL iotk_scan_end(u, 'PP_HEADER')
|
|
|
|
!CALL debug_pseudo_upf(upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_header
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_mesh(u, upf, grid)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
USE radial_grids, ONLY: allocate_radial_grid
|
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
TYPE(radial_grid_type),OPTIONAL,INTENT(INOUT),TARGET :: grid
|
|
|
|
!
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
LOGICAL :: found
|
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_MESH', attr=attr)
|
|
|
|
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'dx', upf%dx, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'mesh', upf%mesh, default=upf%mesh)
|
|
|
|
CALL iotk_scan_attr(attr, 'xmin', upf%xmin, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'rmax', upf%rmax, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'zmesh',upf%zmesh, default=0._dp)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF (present(grid)) THEN
|
|
|
|
CALL allocate_radial_grid(grid, upf%mesh)
|
|
|
|
!
|
|
|
|
grid%dx = upf%dx
|
|
|
|
grid%mesh = upf%mesh
|
|
|
|
grid%xmin = upf%xmin
|
|
|
|
grid%rmax = upf%rmax
|
|
|
|
grid%zmesh = upf%zmesh
|
|
|
|
!
|
|
|
|
upf%grid => grid
|
|
|
|
upf%r => upf%grid%r
|
|
|
|
upf%rab => upf%grid%rab
|
|
|
|
ELSE
|
|
|
|
ALLOCATE( upf%r( upf%mesh ), upf%rab( upf%mesh ) )
|
|
|
|
ENDIF
|
|
|
|
!
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_dat(u, 'PP_R', upf%r(1:upf%mesh))
|
|
|
|
CALL iotk_scan_dat(u, 'PP_RAB', upf%rab(1:upf%mesh))
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
IF (present(grid)) THEN
|
|
|
|
! Reconstruct additional grids
|
|
|
|
upf%grid%r2 = upf%r**2
|
|
|
|
upf%grid%sqr = sqrt(upf%r)
|
2008-04-06 22:50:21 +08:00
|
|
|
upf%grid%rm1 = upf%r**(-1)
|
|
|
|
upf%grid%rm2 = upf%r**(-2)
|
|
|
|
upf%grid%rm3 = upf%r**(-3)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDIF
|
|
|
|
|
|
|
|
CALL iotk_scan_end(u, 'PP_MESH')
|
|
|
|
!
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_mesh
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_nonlocal(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
!
|
|
|
|
INTEGER :: nb,mb,ln,lm,l,nmb,ierr=0
|
|
|
|
!INTEGER :: nb_=-1,mb_=-1,l_=-1,nmb_=-1
|
|
|
|
REAL(DP):: zeros(upf%mesh)
|
|
|
|
LOGICAL :: isnull, found
|
2008-04-16 21:33:43 +08:00
|
|
|
zeros=0._dp
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2009-05-05 15:11:40 +08:00
|
|
|
! modified by AF
|
|
|
|
!IF (upf%tcoulombp) RETURN
|
|
|
|
IF (upf%tcoulombp) upf%nbeta = 0
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Allocate space for non-local part
|
2009-05-05 15:11:40 +08:00
|
|
|
IF ( upf%nbeta == 0) then
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
upf%nqf = 0
|
|
|
|
upf%nqlc= 0
|
|
|
|
upf%qqq_eps= -1._dp
|
|
|
|
upf%kkbeta = 0
|
|
|
|
ALLOCATE( upf%kbeta(1), &
|
|
|
|
upf%lll(1), &
|
|
|
|
upf%beta(upf%mesh,1), &
|
|
|
|
upf%dion(1,1), &
|
|
|
|
upf%rinner(1), &
|
|
|
|
upf%qqq(1,1), &
|
|
|
|
upf%qfunc(upf%mesh,1),&
|
|
|
|
upf%qfcoef(1,1,1,1), &
|
|
|
|
upf%rcut(1), &
|
|
|
|
upf%rcutus(1), &
|
|
|
|
upf%els_beta(1) )
|
2009-05-05 15:11:40 +08:00
|
|
|
! <AF>
|
|
|
|
!CALL iotk_scan_end(u, 'PP_NONLOCAL')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
RETURN
|
2009-05-05 15:11:40 +08:00
|
|
|
END IF
|
|
|
|
!
|
|
|
|
! <AF>
|
|
|
|
CALL iotk_scan_begin(u, 'PP_NONLOCAL')
|
|
|
|
!
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ALLOCATE( upf%kbeta(upf%nbeta), &
|
|
|
|
upf%lll(upf%nbeta), &
|
|
|
|
upf%beta(upf%mesh, upf%nbeta), &
|
|
|
|
upf%dion(upf%nbeta, upf%nbeta),&
|
|
|
|
upf%rcut(upf%nbeta), &
|
|
|
|
upf%rcutus(upf%nbeta), &
|
|
|
|
upf%els_beta(upf%nbeta) )
|
2009-05-05 15:11:40 +08:00
|
|
|
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Read the projectors:
|
|
|
|
DO nb = 1,upf%nbeta
|
|
|
|
CALL iotk_scan_dat(u, 'PP_BETA'//iotk_index( nb ), &
|
|
|
|
upf%beta(:,nb), attr=attr)
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'label', upf%els_beta(nb), default='Xn')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'angular_momentum', upf%lll(nb))
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'cutoff_radius_index', upf%kbeta(nb), default=upf%mesh)
|
|
|
|
CALL iotk_scan_attr(attr, 'cutoff_radius', upf%rcut(nb), default=0._dp)
|
2011-03-15 21:42:15 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'ultrasoft_cutoff_radius', upf%rcutus(nb), default=0._dp)
|
|
|
|
!
|
|
|
|
! Old version of UPF PPs v.2 contained an error in the tag.
|
|
|
|
! To be able to read the old PPs we need the following
|
|
|
|
!
|
|
|
|
IF ( upf%rcutus(nb)==0._DP) &
|
|
|
|
CALL iotk_scan_attr(attr,'norm_conserving_radius',upf%rcutus(nb), &
|
|
|
|
default=0._dp)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
! Read the hamiltonian terms D_ij
|
|
|
|
CALL iotk_scan_dat(u, 'PP_DIJ', upf%dion, attr=attr)
|
2008-04-04 17:12:30 +08:00
|
|
|
! CALL iotk_scan_attr(attr, 'non_zero_elements', upf%nd)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Read the augmentation charge section
|
|
|
|
augmentation : &
|
|
|
|
IF(upf%tvanp .or. upf%tpawp) THEN
|
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_AUGMENTATION', attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'q_with_l', upf%q_with_l)
|
|
|
|
CALL iotk_scan_attr(attr, 'nqf', upf%nqf)
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'nqlc', upf%nqlc, default=2*upf%lmax+1)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF (upf%tpawp) THEN
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr,'shape', upf%paw%augshape, default='UNKNOWN')
|
|
|
|
CALL iotk_scan_attr(attr,'cutoff_r', upf%paw%raug, default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr,'cutoff_r_index', upf%paw%iraug, default=upf%mesh)
|
|
|
|
CALL iotk_scan_attr(attr,'l_max_aug', upf%paw%lmax_aug, default=upf%lmax_rho)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDIF
|
2008-04-09 22:51:58 +08:00
|
|
|
! a negative number means that all qfunc are stored
|
|
|
|
CALL iotk_scan_attr(attr,'augmentation_epsilon',upf%qqq_eps, default=-1._dp)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
ALLOCATE( upf%rinner( upf%nqlc ) )
|
|
|
|
ALLOCATE( upf%qqq ( upf%nbeta, upf%nbeta ) )
|
|
|
|
IF ( upf%q_with_l ) THEN
|
|
|
|
ALLOCATE( upf%qfuncl ( upf%mesh, upf%nbeta*(upf%nbeta+1)/2, 0:2*upf%lmax ) )
|
2009-02-12 22:53:45 +08:00
|
|
|
upf%qfuncl=0._dp
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ELSE
|
|
|
|
ALLOCATE( upf%qfunc (upf%mesh, upf%nbeta*(upf%nbeta+1)/2) )
|
|
|
|
ENDIF
|
|
|
|
!
|
|
|
|
! Read the integrals of the Q functions
|
|
|
|
CALL iotk_scan_dat(u, 'PP_Q',upf%qqq )
|
|
|
|
!
|
|
|
|
! read charge multipoles (only if PAW)
|
2008-04-09 22:51:58 +08:00
|
|
|
IF( upf%tpawp ) THEN
|
|
|
|
ALLOCATE(upf%paw%augmom(upf%nbeta,upf%nbeta, 0:2*upf%lmax))
|
|
|
|
CALL iotk_scan_dat(u, 'PP_MULTIPOLES', upf%paw%augmom)
|
|
|
|
ENDIF
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Read polinomial coefficients for Q_ij expansion at small radius
|
|
|
|
IF(upf%nqf <= 0) THEN
|
|
|
|
upf%rinner(:) = 0._dp
|
|
|
|
ALLOCATE( upf%qfcoef(1,1,1,1) )
|
|
|
|
upf%qfcoef = 0._dp
|
|
|
|
ELSE
|
|
|
|
ALLOCATE( upf%qfcoef( MAX( upf%nqf,1 ), upf%nqlc, upf%nbeta, upf%nbeta ) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_QFCOEF',upf%qfcoef, attr=attr)
|
|
|
|
CALL iotk_scan_dat(u, 'PP_RINNER',upf%rinner, attr=attr)
|
|
|
|
ENDIF
|
|
|
|
!
|
|
|
|
! Read augmentation charge Q_ij
|
|
|
|
ultrasoft_or_paw : &
|
|
|
|
IF( upf%tvanp) THEN
|
|
|
|
DO nb = 1,upf%nbeta
|
|
|
|
ln = upf%lll(nb)
|
|
|
|
DO mb = nb,upf%nbeta
|
|
|
|
lm = upf%lll(mb)
|
|
|
|
nmb = mb * (mb-1) /2 + nb
|
|
|
|
q_with_l : &
|
|
|
|
IF( upf%q_with_l ) THEN
|
|
|
|
DO l = abs(ln-lm),ln+lm,2 ! only even terms
|
|
|
|
CALL iotk_scan_dat(u, 'PP_QIJL'//iotk_index((/nb,mb,l/)),&
|
2008-04-09 22:51:58 +08:00
|
|
|
upf%qfuncl(:,nmb,l),default=zeros,attr=attr)
|
2010-03-13 18:45:42 +08:00
|
|
|
IF( upf%tpawp) upf%qfuncl(upf%paw%iraug+1:,nmb,l) = 0._dp
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDDO
|
|
|
|
ELSE q_with_l
|
|
|
|
CALL iotk_scan_dat(u, 'PP_QIJ'//iotk_index((/nb,mb/)),&
|
2008-04-09 22:51:58 +08:00
|
|
|
upf%qfunc(:,nmb),attr=attr,default=zeros)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDIF q_with_l
|
|
|
|
ENDDO
|
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
ENDIF ultrasoft_or_paw
|
|
|
|
!
|
|
|
|
CALL iotk_scan_end(u, 'PP_AUGMENTATION')
|
|
|
|
!
|
|
|
|
ENDIF augmentation
|
|
|
|
!
|
|
|
|
! Maximum radius of beta projector: outer radius to integrate
|
|
|
|
upf%kkbeta = MAXVAL(upf%kbeta(1:upf%nbeta))
|
2009-01-16 00:34:59 +08:00
|
|
|
! For PAW augmentation charge may extend a bit further:
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IF(upf%tpawp) upf%kkbeta = MAX(upf%kkbeta, upf%paw%iraug)
|
|
|
|
!
|
|
|
|
CALL iotk_scan_end(u, 'PP_NONLOCAL')
|
|
|
|
!
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_nonlocal
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_pswfc(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
!
|
|
|
|
INTEGER :: nw
|
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_PSWFC')
|
|
|
|
!
|
|
|
|
ALLOCATE( upf%chi(upf%mesh,upf%nwfc) )
|
2008-04-04 22:01:35 +08:00
|
|
|
ALLOCATE( upf%els(upf%nwfc), &
|
|
|
|
upf%oc(upf%nwfc), &
|
|
|
|
upf%lchi(upf%nwfc), &
|
|
|
|
upf%nchi(upf%nwfc), &
|
|
|
|
upf%rcut_chi(upf%nwfc), &
|
|
|
|
upf%rcutus_chi(upf%nwfc), &
|
|
|
|
upf%epseu(upf%nwfc) &
|
|
|
|
)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
DO nw = 1,upf%nwfc
|
|
|
|
CALL iotk_scan_dat(u, 'PP_CHI'//iotk_index(nw), &
|
|
|
|
upf%chi(:,nw), attr=attr)
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'label', upf%els(nw), default='Xn')
|
2008-04-04 22:01:35 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'l', upf%lchi(nw))
|
|
|
|
CALL iotk_scan_attr(attr, 'occupation', upf%oc(nw))
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'n', upf%nchi(nw), default=upf%lchi(nw)-1)
|
|
|
|
CALL iotk_scan_attr(attr, 'pseudo_energy', upf%epseu(nw), default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'cutoff_radius', upf%rcut_chi(nw),default=0._dp)
|
|
|
|
CALL iotk_scan_attr(attr, 'ultrasoft_cutoff_radius', upf%rcutus_chi(nw),default=0._dp)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
CALL iotk_scan_end(u, 'PP_PSWFC')
|
|
|
|
!
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_pswfc
|
2008-04-16 18:12:11 +08:00
|
|
|
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_full_wfc(u, upf)
|
2008-04-16 18:12:11 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
2010-03-13 18:45:42 +08:00
|
|
|
LOGICAL :: exst
|
2008-04-16 18:12:11 +08:00
|
|
|
!
|
|
|
|
INTEGER :: nb
|
|
|
|
!
|
|
|
|
IF(.not. upf%has_wfc) RETURN
|
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_FULL_WFC')
|
|
|
|
!
|
|
|
|
ALLOCATE( upf%aewfc(upf%mesh, upf%nbeta) )
|
|
|
|
DO nb = 1,upf%nbeta
|
|
|
|
CALL iotk_scan_dat(u, 'PP_AEWFC'//iotk_index(nb), &
|
|
|
|
upf%aewfc(:,nb), attr=attr)
|
|
|
|
ENDDO
|
|
|
|
|
2010-03-13 18:45:42 +08:00
|
|
|
IF (upf%has_so .and. upf%tpawp) THEN
|
|
|
|
ALLOCATE( upf%paw%aewfc_rel(upf%mesh, upf%nbeta) )
|
|
|
|
nb_loop: DO nb = 1,upf%nbeta
|
|
|
|
CALL iotk_scan_dat(u, 'PP_AEWFC_REL'//iotk_index(nb), &
|
|
|
|
upf%paw%aewfc_rel(:,nb), attr=attr, found=exst)
|
|
|
|
IF (.not.exst) THEN
|
|
|
|
upf%paw%aewfc_rel=0.0_DP
|
|
|
|
EXIT nb_loop
|
|
|
|
ENDIF
|
|
|
|
ENDDO nb_loop
|
|
|
|
ENDIF
|
|
|
|
|
2008-04-16 18:12:11 +08:00
|
|
|
ALLOCATE( upf%pswfc(upf%mesh, upf%nbeta) )
|
|
|
|
DO nb = 1,upf%nbeta
|
|
|
|
CALL iotk_scan_dat(u, 'PP_PSWFC'//iotk_index(nb), &
|
|
|
|
upf%pswfc(:,nb), attr=attr)
|
|
|
|
ENDDO
|
|
|
|
CALL iotk_scan_end(u, 'PP_FULL_WFC')
|
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_full_wfc
|
2008-04-16 18:12:11 +08:00
|
|
|
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_spin_orb(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
!
|
|
|
|
INTEGER :: nw, nb
|
|
|
|
!
|
|
|
|
IF (.not. upf%has_so) RETURN
|
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_SPIN_ORB')
|
|
|
|
!
|
2009-01-16 00:34:59 +08:00
|
|
|
ALLOCATE (upf%nn(upf%nwfc))
|
|
|
|
ALLOCATE (upf%jchi(upf%nwfc))
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
DO nw = 1,upf%nwfc
|
|
|
|
CALL iotk_scan_empty(u, 'PP_RELWFC'//iotk_index(nw),&
|
|
|
|
attr=attr)
|
2008-04-09 22:51:58 +08:00
|
|
|
!CALL iotk_scan_attr(attr, 'els', upf%els(nw)) ! already read
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'nn', upf%nn(nw))
|
2008-04-09 22:51:58 +08:00
|
|
|
!CALL iotk_scan_attr(attr, 'lchi', upf%lchi(nw)) ! already read
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'jchi', upf%jchi(nw))
|
2008-04-09 22:51:58 +08:00
|
|
|
!CALL iotk_scan_attr(attr, 'oc', upf%oc(nw)) ! already read
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDDO
|
|
|
|
!
|
2009-01-16 00:34:59 +08:00
|
|
|
ALLOCATE(upf%jjj(upf%nbeta))
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
DO nb = 1,upf%nbeta
|
|
|
|
CALL iotk_scan_empty(u, 'PP_RELBETA'//iotk_index(nb),&
|
|
|
|
attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'lll', upf%lll(nb))
|
|
|
|
CALL iotk_scan_attr(attr, 'jjj', upf%jjj(nb))
|
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
CALL iotk_scan_end(u, 'PP_SPIN_ORB')
|
|
|
|
!
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_spin_orb
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_paw(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
|
|
|
|
!
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
!
|
|
|
|
INTEGER :: nb,nb1
|
|
|
|
|
|
|
|
IF (.not. upf%tpawp ) RETURN
|
|
|
|
|
|
|
|
CALL iotk_scan_begin(u, 'PP_PAW', attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'paw_data_format', upf%paw_data_format)
|
2008-04-04 17:12:30 +08:00
|
|
|
IF(upf%paw_data_format /= 2) &
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL errore('read_upf_v2::paw',&
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
'Unknown format of PAW data.',1)
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'core_energy', upf%paw%core_energy, default=0._dp)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Full occupation (not only > 0 ones)
|
|
|
|
ALLOCATE( upf%paw%oc(upf%nbeta) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_OCCUPATIONS',upf%paw%oc)
|
|
|
|
!
|
|
|
|
! All-electron core charge
|
|
|
|
ALLOCATE( upf%paw%ae_rho_atc(upf%mesh) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_AE_NLCC', upf%paw%ae_rho_atc)
|
|
|
|
!
|
|
|
|
! All-electron local potential
|
|
|
|
ALLOCATE( upf%paw%ae_vloc(upf%mesh) )
|
|
|
|
CALL iotk_scan_dat(u, 'PP_AE_VLOC', upf%paw%ae_vloc)
|
|
|
|
!
|
|
|
|
ALLOCATE(upf%paw%pfunc(upf%mesh, upf%nbeta,upf%nbeta) )
|
|
|
|
upf%paw%pfunc(:,:,:) = 0._dp
|
2010-03-13 18:45:42 +08:00
|
|
|
IF (upf%has_so) THEN
|
|
|
|
ALLOCATE(upf%paw%pfunc_rel(upf%mesh, upf%nbeta,upf%nbeta) )
|
|
|
|
upf%paw%pfunc_rel(:,:,:) = 0._dp
|
|
|
|
ENDIF
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
DO nb=1,upf%nbeta
|
|
|
|
DO nb1=1,nb
|
|
|
|
upf%paw%pfunc (1:upf%mesh, nb, nb1) = &
|
|
|
|
upf%aewfc(1:upf%mesh, nb) * upf%aewfc(1:upf%mesh, nb1)
|
2010-03-13 18:45:42 +08:00
|
|
|
IF (upf%has_so) THEN
|
|
|
|
upf%paw%pfunc_rel (1:upf%paw%iraug, nb, nb1) = &
|
|
|
|
upf%paw%aewfc_rel(1:upf%paw%iraug, nb) * &
|
|
|
|
upf%paw%aewfc_rel(1:upf%paw%iraug, nb1)
|
|
|
|
!
|
|
|
|
! The small component is added to pfunc. pfunc_rel is useful only
|
|
|
|
! to add a small magnetic contribution
|
|
|
|
!
|
|
|
|
upf%paw%pfunc (1:upf%paw%iraug, nb, nb1) = &
|
|
|
|
upf%paw%pfunc (1:upf%paw%iraug, nb, nb1) + &
|
|
|
|
upf%paw%pfunc_rel (1:upf%paw%iraug, nb, nb1)
|
|
|
|
ENDIF
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
upf%paw%pfunc(upf%paw%iraug+1:,nb,nb1) = 0._dp
|
|
|
|
!
|
|
|
|
upf%paw%pfunc (1:upf%mesh, nb1, nb) = upf%paw%pfunc (1:upf%mesh, nb, nb1)
|
2010-03-13 18:45:42 +08:00
|
|
|
IF (upf%has_so) upf%paw%pfunc_rel (1:upf%mesh, nb1, nb) = &
|
|
|
|
upf%paw%pfunc_rel (1:upf%mesh, nb, nb1)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ENDDO
|
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
! Pseudo wavefunctions (not only the ones for oc > 0)
|
2008-04-16 18:12:11 +08:00
|
|
|
! All-electron wavefunctions
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ALLOCATE(upf%paw%ptfunc(upf%mesh, upf%nbeta,upf%nbeta) )
|
|
|
|
upf%paw%ptfunc(:,:,:) = 0._dp
|
|
|
|
DO nb=1,upf%nbeta
|
|
|
|
DO nb1=1,upf%nbeta
|
|
|
|
upf%paw%ptfunc (1:upf%mesh, nb, nb1) = &
|
|
|
|
upf%pswfc(1:upf%mesh, nb) * upf%pswfc(1:upf%mesh, nb1)
|
|
|
|
upf%paw%ptfunc(upf%paw%iraug+1:,nb,nb1) = 0._dp
|
|
|
|
!
|
|
|
|
upf%paw%ptfunc (1:upf%mesh, nb1, nb) = upf%paw%ptfunc (1:upf%mesh, nb, nb1)
|
|
|
|
ENDDO
|
|
|
|
ENDDO
|
|
|
|
!
|
|
|
|
! Finalize
|
|
|
|
CALL iotk_scan_end(u, 'PP_PAW')
|
|
|
|
|
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_paw
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2012-02-14 04:29:47 +08:00
|
|
|
SUBROUTINE read_upf_gipaw(u, upf)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
IMPLICIT NONE
|
|
|
|
INTEGER,INTENT(IN) :: u ! i/o unit
|
|
|
|
TYPE(pseudo_upf),INTENT(INOUT) :: upf ! the pseudo data
|
|
|
|
INTEGER :: ierr ! /= 0 if something went wrong
|
|
|
|
!
|
|
|
|
CHARACTER(len=iotk_attlenx) :: attr
|
|
|
|
!
|
|
|
|
INTEGER :: nb
|
|
|
|
IF (.not. upf%has_gipaw ) RETURN
|
2010-07-06 20:04:34 +08:00
|
|
|
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_begin(u, 'PP_GIPAW', attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'gipaw_data_format', upf%gipaw_data_format)
|
|
|
|
IF(upf%gipaw_data_format /= 2) &
|
2012-02-14 04:29:47 +08:00
|
|
|
CALL infomsg('read_upf_v2::gipaw','Unknown format version')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
CALL iotk_scan_begin(u, 'PP_GIPAW_CORE_ORBITALS', attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'number_of_core_orbitals', upf%gipaw_ncore_orbitals)
|
|
|
|
ALLOCATE ( upf%gipaw_core_orbital_n(upf%gipaw_ncore_orbitals) )
|
2008-04-09 22:51:58 +08:00
|
|
|
ALLOCATE ( upf%gipaw_core_orbital_el(upf%gipaw_ncore_orbitals) )
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
ALLOCATE ( upf%gipaw_core_orbital_l(upf%gipaw_ncore_orbitals) )
|
|
|
|
ALLOCATE ( upf%gipaw_core_orbital(upf%mesh,upf%gipaw_ncore_orbitals) )
|
|
|
|
DO nb = 1,upf%gipaw_ncore_orbitals
|
|
|
|
CALL iotk_scan_dat(u, 'PP_GIPAW_CORE_ORBITAL'//iotk_index(nb), &
|
|
|
|
upf%gipaw_core_orbital(:,nb), attr=attr)
|
|
|
|
CALL iotk_scan_attr(attr, 'label', upf%gipaw_core_orbital_el(nb))
|
|
|
|
CALL iotk_scan_attr(attr, 'n', upf%gipaw_core_orbital_n(nb))
|
|
|
|
CALL iotk_scan_attr(attr, 'l', upf%gipaw_core_orbital_l(nb))
|
|
|
|
ENDDO
|
|
|
|
CALL iotk_scan_end(u, 'PP_GIPAW_CORE_ORBITALS')
|
2010-07-06 20:04:34 +08:00
|
|
|
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
! Read valence all-electron and pseudo orbitals and their labels
|
2012-11-27 19:39:10 +08:00
|
|
|
!
|
2010-07-06 20:04:34 +08:00
|
|
|
IF (upf%paw_as_gipaw) THEN
|
|
|
|
!READ PAW DATA INSTEAD OF GIPAW
|
|
|
|
upf%gipaw_wfs_nchannels = upf%nbeta
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_el(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ll(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcut(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcutus(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ae(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ps(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
upf%gipaw_wfs_el(nb) = upf%els_beta(nb)
|
|
|
|
upf%gipaw_wfs_ll(nb) = upf%lll(nb)
|
|
|
|
upf%gipaw_wfs_ae(:,nb) = upf%aewfc(:,nb)
|
|
|
|
ENDDO
|
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
upf%gipaw_wfs_ps(:,nb) = upf%pswfc(:,nb)
|
|
|
|
ENDDO
|
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ae(upf%mesh) )
|
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ps(upf%mesh) )
|
2012-11-27 19:39:10 +08:00
|
|
|
upf%gipaw_vlocal_ae(:)= upf%paw%ae_vloc(:)
|
|
|
|
upf%gipaw_vlocal_ps(:)= upf%vloc(:)
|
2010-07-06 20:04:34 +08:00
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
upf%gipaw_wfs_rcut(nb)=upf%rcut(nb)
|
|
|
|
upf%gipaw_wfs_rcutus(nb)=upf%rcutus(nb)
|
|
|
|
ENDDO
|
|
|
|
ELSEIF (upf%tcoulombp) THEN
|
|
|
|
upf%gipaw_wfs_nchannels = 1
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_el(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ll(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcut(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcutus(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ae(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ps(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
upf%gipaw_wfs_el(nb) = "1S"
|
|
|
|
upf%gipaw_wfs_ll(nb) = 0
|
|
|
|
upf%gipaw_wfs_ae(:,nb) = 0.0d0
|
|
|
|
upf%gipaw_wfs_ps(:,nb) = 0.0d0
|
|
|
|
ENDDO
|
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ae(upf%mesh) )
|
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ps(upf%mesh) )
|
|
|
|
upf%gipaw_vlocal_ae(:)= 0.0d0
|
|
|
|
upf%gipaw_vlocal_ps(:)= 0.0d0
|
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
upf%gipaw_wfs_rcut(nb)=1.0d0
|
|
|
|
upf%gipaw_wfs_rcutus(nb)=1.0d0
|
|
|
|
ENDDO
|
|
|
|
ELSE
|
|
|
|
CALL iotk_scan_begin(u, 'PP_GIPAW_ORBITALS', attr=attr)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'number_of_valence_orbitals', upf%gipaw_wfs_nchannels)
|
2010-07-06 20:04:34 +08:00
|
|
|
ALLOCATE ( upf%gipaw_wfs_el(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ll(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcut(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_rcutus(upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ae(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
ALLOCATE ( upf%gipaw_wfs_ps(upf%mesh,upf%gipaw_wfs_nchannels) )
|
|
|
|
|
|
|
|
DO nb = 1,upf%gipaw_wfs_nchannels
|
|
|
|
CALL iotk_scan_begin(u, 'PP_GIPAW_ORBITAL'//iotk_index(nb), attr=attr)
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'label', upf%gipaw_wfs_el(nb))
|
|
|
|
CALL iotk_scan_attr(attr, 'l', upf%gipaw_wfs_ll(nb))
|
|
|
|
CALL iotk_scan_attr(attr, 'cutoff_radius', upf%gipaw_wfs_rcut(nb))
|
2008-04-09 22:51:58 +08:00
|
|
|
CALL iotk_scan_attr(attr, 'ultrasoft_cutoff_radius', upf%gipaw_wfs_rcutus(nb),&
|
|
|
|
default=upf%gipaw_wfs_rcut(nb))
|
2010-07-06 20:04:34 +08:00
|
|
|
! read all-electron orbital
|
|
|
|
CALL iotk_scan_dat(u, 'PP_GIPAW_WFS_AE', upf%gipaw_wfs_ae(:,nb))
|
|
|
|
! read pseudo orbital
|
|
|
|
CALL iotk_scan_dat(u, 'PP_GIPAW_WFS_PS', upf%gipaw_wfs_ps(:,nb))
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
2010-07-06 20:04:34 +08:00
|
|
|
CALL iotk_scan_end(u, 'PP_GIPAW_ORBITAL'//iotk_index(nb))
|
|
|
|
ENDDO
|
|
|
|
CALL iotk_scan_end(u, 'PP_GIPAW_ORBITALS')
|
|
|
|
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
2009-01-16 00:34:59 +08:00
|
|
|
! Read all-electron and pseudo local potentials
|
2010-07-06 20:04:34 +08:00
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ae(upf%mesh) )
|
|
|
|
ALLOCATE ( upf%gipaw_vlocal_ps(upf%mesh) )
|
|
|
|
CALL iotk_scan_begin(u, 'PP_GIPAW_VLOCAL')
|
|
|
|
CALL iotk_scan_dat(u, 'PP_GIPAW_VLOCAL_AE',upf%gipaw_vlocal_ae(:))
|
2012-11-27 21:57:46 +08:00
|
|
|
CALL iotk_scan_dat(u, 'PP_GIPAW_VLOCAL_PS',upf%gipaw_vlocal_ps(:))
|
2010-07-06 20:04:34 +08:00
|
|
|
CALL iotk_scan_end(u, 'PP_GIPAW_VLOCAL')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
|
2010-07-06 20:04:34 +08:00
|
|
|
ENDIF
|
|
|
|
CALL iotk_scan_end(u, 'PP_GIPAW')
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
RETURN
|
2012-02-14 04:29:47 +08:00
|
|
|
END SUBROUTINE read_upf_gipaw
|
This is a quite complex check-in, but actually not very much is done. Changelog follows.
LP
UPF file format updated completely, UPFv2 introduced:
* ld1.x can still produce old format, with the switch upf_v1_format=.true. in inputp
this is disabled by default, but we can discuss if it should be the opposite.
* pw.x cp.x and all utilities should notice no difference
* some utilities in upftools still need to be updated, anyway conversion UPFv1 to UPFv2
is very easy, so this should be no big issue
* starting from now to produce an UPF file you need to fill the pseudo_upf derivedd type
and feed it to write_upf woutine in upf_module (Modules/upf.f90)
* extensive use of iotk
I have tried to make the new format as self contained as possible, e.g. there should be
minimal need for post-processing after the data is read, no more reconstruction of known
quantities, and no more odd syntax to save negligible quantity of space. Also the human
readable section is a bit richer, all the rest is more machine readable.
I hope this will not cause any throuble, and tried really hard to, all examples and all
tests works as fine as before and gives (what really looks like) the same results.
Other changes that I needed to make:
* radial grids are now allocatable, they management is a bit less of a hack too
* paw and uspp augmentation are stored in the same place
* paw print total all-electron energy if all atoms are paw, not very useful, but nice
* most of the pseudopotential-writing reading files have been renamed to some more
logical name, I spare you the list. E.g. read_oldpseudo -> read_pseudo_rrkj3
* paw_t derived type was only used in atomic, so I have put it there (as the pseudo_type
module take ages to recompile it was awkward to leave it there).
PAW tests inserted in test/ there are 6 of them, as a consequence I have also put 5 paw
pseudopotentials in the pseudo/ directory.
I will update the PAW scf examples soon, by deleting them (as running a pw with a PAW
pseudopotential requires no option at all). PAW generation examples should be updated.
A lot of small bugfixes here & there mostly uninitialized variables or unallocated
pointers used as subrotuine arguments.
git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@4769 c92efa57-630b-4861-b058-cf58834340f0
2008-04-03 23:50:43 +08:00
|
|
|
!
|
|
|
|
END SUBROUTINE read_upf_v2
|
|
|
|
!
|
|
|
|
END MODULE read_upf_v2_module
|