Bug fix: Now fix parsing of categories more than 2 indent levels prior

It was discovered that this YAML markup was not being parsed properly
in qfyaml 0.3.2:

   operations:
     transport:
       passive_species:
         CH3ITracer:
           long_name: Methyl_iodide
           mol_wt_in_g: 142.0
           lifetime_in_s: 4.32e5
           default_bkg_conc_in_vv: 1.0e-20
     wet_deposition:
       activate: true

The wet_deposition category was given the wrong indentation depth,
and as such, operations:wet_deposition:activate was parsed as FALSE
but should have been TRUE.

This issue is now fixed and we will release this as qfyaml 0.3.0.

NOTE: QFYAML works best if a constant indentation (e.g. always 2 spaces
for each level) is used.  You can use an editor such as Emacs to format
your YAML input files accordingly.

Also updated the test programs to test this behavior.

Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>

docs/source/known-bugs.rst

@@ -6,6 +6,36 @@ This page lists known bugs in :program:`qfyaml`. See the `Github
 issues <http:s://github.com/yantosca/qfyaml/issues>`_ page for updates
 on their status.
 
+*************
+Version 0.3.2
+*************
+
+Error parsing categories
+========================
+
+We discovered an error parsing this YAML file, where the
+:code:`wet_deposition` tag is more than 2 indentation levels behind
+behind the previous line.
+
+.. code-block:: yaml
+
+   operations:
+     transport:
+       passive_species:
+         CH3ITracer:
+           long_name: Methyl_iodide
+           mol_wt_in_g: 142.0
+           lifetime_in_s: 4.32e5
+           default_bkg_conc_in_vv: 1.0e-20
+     wet_deposition:
+       activate: true
+
+This has now been fixed in qfyaml 0.3.3.
+
+NOTE: For best results with qfyaml, we recommend formatting YAML files
+so that they contain a consistent indentation level throughout the
+file (i.e. such as 2 or 4 spaces).  Editors such as Emacs can do this easily.
+
 *************
 Version 0.3.0
 *************

src/qfyaml_mod.F90

@@ -40,6 +40,7 @@ MODULE QFYAML_Mod
   PUBLIC :: QFYAML_CleanUp
   PUBLIC :: QFYAML_Get
   PUBLIC :: QFYAML_Check
+  PUBLIC :: QFYAML_FindDepth
   PUBLIC :: QFYAML_FindNextHigher
   PUBLIC :: QFYAML_Init
   PUBLIC :: QFYAML_Merge
@@ -741,7 +742,8 @@ SUBROUTINE Parse_Line( yml,        yml_anchored, set_by,                   &
     ! SAVEd variables
     LOGICAL,                      SAVE :: is_list_var                = .FALSE.
     INTEGER,                      SAVE :: last_pos                   = 0
-    INTEGER,                      SAVE :: cat_pos(20)                = 0
+    INTEGER,                      SAVE :: indent                     = 0
+    INTEGER,                      SAVE :: cat_pos(QFYAML_MaxStack)   = 0
     INTEGER,                      SAVE :: cat_index                  = 0
     CHARACTER(LEN=QFYAML_NamLen), SAVE :: cat_stack(QFYAML_MaxStack) = ""
 
@@ -805,23 +807,27 @@ SUBROUTINE Parse_Line( yml,        yml_anchored, set_by,                   &
           ! If this category starts further along the line than the last
           ! category, then increment index and add its position to the stack.
           IF ( pos > last_pos ) THEN
+             indent             = last_pos - pos
              cat_index          = cat_index + 1
              cat_pos(cat_index) = pos
           ENDIF
 
           ! If this category starts earlier along the line than the last
           ! category, then decrement index and add its position to the stack.
-          ! last category, then this
+          ! NOTE: This algorithm will work best if we assume a constant
+          ! indentation level.  Best to use an editor such as emacs
+          ! to enforce a consistent indentation throughout the file.
           IF ( pos < last_pos ) THEN
-             cat_index          = cat_index - 1
+             indent    = last_pos - pos
+             cat_index = cat_index - ( indent / 2 )
              cat_pos(cat_index) = pos
           ENDIF
 
           ! If the index is negative or the category begins at the first
           ! character of the line, then set index to 1 and store its
           ! starting position in the first element of the stack.
           IF ( cat_index <= 0 .or. pos == 1 ) THEN
-             cat_index          = 1
+             cat_index = 1
              cat_pos(cat_index) = pos
           ENDIF
 
@@ -927,9 +933,9 @@ SUBROUTINE Parse_Line( yml,        yml_anchored, set_by,                   &
           category = cat_stack(C)
           IF ( C > 1 ) THEN
              DO CC = C-1, 1, -1
-                category = TRIM( cat_stack(CC) ) // &
-                           QFYAML_Category_Separator // &
-                           TRIM( category )
+                category = TRIM( cat_stack(CC)       )                    // &
+                           QFYAML_Category_Separator                      // &
+                           TRIM( category            )
              ENDDO
           ENDIF
           EXIT
@@ -941,16 +947,15 @@ SUBROUTINE Parse_Line( yml,        yml_anchored, set_by,                   &
           category = cat_stack( MAX( C-1, 1 ) )
           IF ( C-1 > 1 ) THEN
              DO CC = C-2, 1, -1
-                category = TRIM( cat_stack(CC) ) // &
-                           QFYAML_Category_Separator // &
-                           TRIM( category )
+                category = TRIM( cat_stack(CC)       )                   // &
+                           QFYAML_Category_Separator                     // &
+                           TRIM( category            )
              ENDDO
           ENDIF
           EXIT
        ENDIF
     ENDDO
 
-
     ! Test if the variable is a YAML anchor
     IF ( var_name == "<<" ) THEN
 
@@ -1498,6 +1503,51 @@ END SUBROUTINE QFYAML_check
 !
 ! !IROUTINE: QFYAML_FindNextHigher
 !
+! !DESCRIPTION: For a given category or variable name, returns its depth
+!  (i.e. indentation level).  This is equal to the number of separator
+!  strings.
+!\\
+!\\
+! !INTERFACE:
+!
+  FUNCTION QFYAML_FindDepth( name ) RESULT( depth )
+!
+! !INPUT PARAMETERS:
+!
+    CHARACTER(LEN=*), INTENT(IN)  :: name
+!
+! RETURN VALUE:
+!
+    INTEGER                       :: depth
+!EOP
+!------------------------------------------------------------------------------
+!BOC
+!
+! !LOCAL VARIABLES:
+!
+    INTEGER :: ix, c
+
+    ! Keep searching for all category separators
+    ! until there aren't any more.
+    depth = 1
+    c     = LEN_TRIM( name )
+    DO
+       ix = INDEX( name(1:c), QFYAML_Category_Separator, back=.TRUE. )
+       IF ( ix <= 1 ) EXIT
+       depth = depth + 1
+       c = ix - 1
+    ENDDO
+
+  END FUNCTION QFYAML_FindDepth
+!EOC
+!------------------------------------------------------------------------------
+! QFYAML: Bob Yantosca | yantosca@seas.harvard.edu | Apr 2020
+! Based on existing package https://github.com/jannisteunissen/config_fortran
+!------------------------------------------------------------------------------
+!BOP
+!
+! !IROUTINE: QFYAML_FindNextHigher
+!
 ! !DESCRIPTION: Finds variables that are one category depth higher than
 !  a given target string (trg_str).  Returns the number of variables that
 !  match this criteria (n_matches), as well as the variables themselves

test/qfyaml.yml

@@ -1,5 +1,5 @@
 ---
-#### input.yml - Input for test_qfyaml.F90
+#### qfyaml.yml - Input for test_qfyaml.F90
 
 #### Test going back & forth between levels
 weather:
@@ -10,7 +10,18 @@ weather:
       units: K
   pressure: 1013.25
 
-#### Test parsing different types of input
+operations:
+  transport:
+    passive_species:
+      CH3ITracer:
+        long_name: Methyl_iodide
+        mol_wt_in_g: 142.0
+        lifetime_in_s: 4.32e5
+        default_bkg_conc_in_vv: 1.0e-20
+  wet_deposition:
+    activate: true
+
+##### Test parsing different types of input
 author:
   age: 29
   fav_reals: [1.0, 2.0]
@@ -47,3 +58,5 @@ filename: another_file
 author_name:
   first: Homer
   full: Homer J. Simpson
+
+

test/test_qfyaml.F90

@@ -65,63 +65,68 @@ PROGRAM Test_QFYAML
 
   key   = "author%age"
   v_int = -999
-  CALL QFYAML_Add_Get( yml, key, v_int, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_int, "", RC )
   WRITE( 6, "(a32, "" : "", i7)") ADJUSTL(key), v_int
 
   key    = "author%fav_reals"
   ALLOCATE( a_real(2) )
   a_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, a_real, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), a_real, "", RC )
   WRITE( 6, "(a32, "" : "", 2f7.2)") ADJUSTL(key), a_real
   DEALLOCATE( a_real )
 
   key    = "author%more_reals"
   ALLOCATE( a_real(4) )
   a_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, a_real, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), a_real, "", RC )
   WRITE( 6, "(a32, "" : "", 4f11.6)") ADJUSTL(key), a_real
   DEALLOCATE( a_real )
 
   key    = "author%lots_of_work"
   v_bool = .FALSE.
-  CALL QFYAML_Add_Get( yml, key, v_bool, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_bool, "", RC )
   WRITE( 6, "(a32, "" : "", l7)") ADJUSTL(key), v_bool
 
   key    = "author_name%first"
   v_str  = ""
-  CALL QFYAML_Add_Get( yml, key, v_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_str, "", RC )
   WRITE( 6, "(a32, "" : "", a)") ADJUSTL(key), TRIM(v_str)
 
   key    = "author_name%full"
   v_str  = ""
-  CALL QFYAML_Add_Get( yml, key, v_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_str, "", RC )
   WRITE( 6, "(a32, "" : "", a)") ADJUSTL(key), TRIM(v_str)
 
   key    = "filename"
   v_str = ""
-  CALL QFYAML_Add_Get( yml, key, v_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_str, "", RC )
   WRITE( 6, "(a32, "" : "", a)") ADJUSTL(key), TRIM(v_str)
 
   key    = "weather%humidity"
   v_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, v_real, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_real, "", RC )
   WRITE( 6, "(a32, "" : "", f13.6)") ADJUSTL(key), v_real
 
   key    = "weather%temperature%daily"
   v_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, v_real, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_real, "", RC )
   WRITE( 6, "(a32, "" : "", f13.6)") ADJUSTL(key), v_real
 
   key    = "weather%temperature%weekly%units"
   v_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, v_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_str, "", RC )
   WRITE( 6, "(a32, "" : "", a)") ADJUSTL(key), TRIM(v_str)
 
   key    = "weather%pressure"
   v_real = -999.0_yp
-  CALL QFYAML_Add_Get( yml, key, v_real, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_real, "", RC )
   WRITE( 6, "(a32, "" : "", f13.6)") ADJUSTL(key), v_real
 
+  key    = "operations%wet_deposition%activate"
+  v_bool = .FALSE.
+  CALL QFYAML_Add_Get( yml, TRIM( key ), v_bool, "", RC )
+  WRITE( 6, "(a32, "" : "", l7)") ADJUSTL(key), v_bool
+
   WRITE( 6, '(/, a)' ) '### FIND NEXT-HIGHER VARIABLES IN "weather"'
 
   CALL QFYAML_FindNextHigher( yml, "weather%", match_ct, match_vars )
@@ -135,7 +140,7 @@ PROGRAM Test_QFYAML
   key   = "fruits"
   ALLOCATE( a_str(3) )
   a_str = ""
-  CALL QFYAML_Add_Get( yml, key, a_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), a_str, "", RC )
   WRITE( 6, "(a)" ) TRIM(key)
   DO N = 1, SIZE( a_str )
      print*, N, TRIM( a_str(N) )
@@ -146,7 +151,7 @@ PROGRAM Test_QFYAML
   key   = "more_fruits%p_fruits"
   ALLOCATE( a_str(4) )
   a_str = ""
-  CALL QFYAML_Add_Get( yml, key, a_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), a_str, "", RC )
   WRITE( 6, "(a)" ) TRIM(key)
   DO N = 1, SIZE( a_str )
      print*, N, TRIM( a_str(N) )
@@ -158,7 +163,7 @@ PROGRAM Test_QFYAML
   key   = "even_more_fruits%exotic_fruits%hard_to_find"
   ALLOCATE( a_str(5) )
   a_str = ""
-  CALL QFYAML_Add_Get( yml, key, a_str, "", RC )
+  CALL QFYAML_Add_Get( yml, TRIM( key ), a_str, "", RC )
   WRITE( 6, "(a)" ) TRIM(key)
   DO N = 1, SIZE( a_str )
      print*, N, TRIM( a_str(N) )

test/test_species_database.F90

@@ -33,7 +33,7 @@ PROGRAM Test_Species_Database
 
   ! Strings
   CHARACTER(LEN=14)           :: tag
-  CHARACTER(LEN=14)           :: spc
+  CHARACTER(LEN=31)           :: spc
   CHARACTER(LEN=255)          :: v_str
   CHARACTER(LEN=255)          :: key
   CHARACTER(LEN=255)          :: fileName
@@ -44,7 +44,7 @@ PROGRAM Test_Species_Database
 
   ! String arrays
   CHARACTER(LEN=17)           :: tags(46)
-  CHARACTER(LEN=14)           :: species(11)
+  CHARACTER(LEN=31)          :: species(11)
 
   ! Objects
   TYPE(QFYAML_t)              :: yml
@@ -77,8 +77,8 @@ PROGRAM Test_Species_Database
 
   RC         = QFYAML_SUCCESS
   mw_g       = MISSING_INT
-  species(1) = "ACTA"
-  species(2) = "ALD2"
+  species(1) = "CO"
+  species(2) = "COAnthroEmis25dayTracer"
   species(3) = "ALK4"
   species(4) = "ASOA1"
   species(5) = "ASOA2"