From 8cc459fad24612c4e6f30862a55a70d74a355116 Mon Sep 17 00:00:00 2001
From: Stefan Appelhoff <stefan.appelhoff@mailbox.org>
Date: Tue, 18 Oct 2022 01:40:22 +0200
Subject: [PATCH] CI: implement codespell check for custom word mappings
 (#1322)

* subdir -> sub-dir: use hyphen

* forgot some, now complete

* sub-dir -> subdir, without hyphen

* try codespell without gh action

* fix reqs

* try custom directory

* more codespell tests

* fix config

* folder -> directory, and other fixes

* more fixes

* ignore .git

* fix flake8, do not touch ipynb

* use codespell config and gh action

* Update .codespell_dict

* try feeding a config

* try if config is parsed regardless

* fix again, codespell works

* test plurals

* plurals dont seem to work auto, so add to dict

* add codespell to pre-commit hooks
---
 .codespell_dict                                  |  6 ++++++
 .codespellrc                                     |  6 ++++++
 .github/workflows/validation.yml                 |  3 ---
 .pre-commit-config.yaml                          |  4 ++++
 src/02-common-principles.md                      | 10 +++++-----
 .../01-magnetic-resonance-imaging-data.md        | 16 ++++++++--------
 .../02-magnetoencephalography.md                 | 12 ++++++------
 .../03-electroencephalography.md                 | 10 +++++-----
 .../04-intracranial-electroencephalography.md    | 10 +++++-----
 .../07-behavioral-experiments.md                 |  2 +-
 .../09-positron-emission-tomography.md           |  4 ++--
 src/04-modality-specific-files/10-microscopy.md  |  4 ++--
 src/appendices/file-collections.md               |  4 ++--
 src/schema/objects/extensions.yaml               | 16 ++++++++--------
 tools/schemacode/bidsschematools/render/text.py  |  2 +-
 .../bidsschematools/tests/test_rules.py          |  2 +-
 16 files changed, 62 insertions(+), 49 deletions(-)
 create mode 100644 .codespell_dict
 create mode 100644 .codespellrc

diff --git a/.codespell_dict b/.codespell_dict
new file mode 100644
index 0000000000..81adf22162
--- /dev/null
+++ b/.codespell_dict
@@ -0,0 +1,6 @@
+sub-directory->subdirectory
+sub-directories->subdirectories
+file name->filename
+file names->filenames
+folder->directory
+folders->directories
diff --git a/.codespellrc b/.codespellrc
new file mode 100644
index 0000000000..28aeafee0c
--- /dev/null
+++ b/.codespellrc
@@ -0,0 +1,6 @@
+[codespell]
+skip = *.js,*.svg,*.eps,.git
+ignore-words-list = fo,te,als
+builtin = clear,rare
+dictionary = .codespell_dict
+exclude-file = src/CHANGES.md
diff --git a/.github/workflows/validation.yml b/.github/workflows/validation.yml
index 0492f0190b..91ba4d49e0 100644
--- a/.github/workflows/validation.yml
+++ b/.github/workflows/validation.yml
@@ -18,9 +18,6 @@ jobs:
     steps:
       - uses: actions/checkout@v3
       - uses: codespell-project/actions-codespell@master
-        with:
-          skip: js,*.svg,*.eps
-          ignore_words_list: fo,te,als
 
   # Markdown formatting
   remark:
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 16a07dd2ee..769560ee89 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -12,3 +12,7 @@ repos:
       - id: check-ast
       - id: check-added-large-files
       - id: check-case-conflict
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.2.2
+    hooks:
+      - id: codespell
diff --git a/src/02-common-principles.md b/src/02-common-principles.md
index 4844c1415e..177384bb3f 100644
--- a/src/02-common-principles.md
+++ b/src/02-common-principles.md
@@ -86,7 +86,7 @@ instead of specified with an `n/a` value, or included as an empty file
 
 ## Filesystem structure
 
-Data for each subject are placed in sub-directories named "`sub-<label>`",
+Data for each subject are placed in subdirectories named "`sub-<label>`",
 where string "`<label>`" is substituted with the unique identification
 label of each subject.
 Additional information on each participant MAY be provided in a
@@ -94,17 +94,17 @@ Additional information on each participant MAY be provided in a
 in the root directory of the dataset.
 
 If data for the subject were acquired across multiple sessions, then within
-the subject directory resides sub-directories named "`ses-<label>`",
+the subject directory resides subdirectories named "`ses-<label>`",
 where string "`<label>`" is substituted with a unique identification
 label for each session.
 In datasets where at least one subject has more than one session, this
-additional sub-directory later SHOULD be added for all subjects in the dataset.
+additional subdirectory later SHOULD be added for all subjects in the dataset.
 Additional information on each session MAY be provided in a
 [sessions file](03-modality-agnostic-files.md#sessions-file)
 within the subject directory.
 
-Within the session sub-directory (or the subject sub-directory if no
-session sub-directories are present) are sub-directories named according to
+Within the session subdirectory (or the subject subdirectory if no
+session subdirectories are present) are subdirectories named according to
 data type as defined above.
 A data type directory SHOULD NOT be defined if there are no files to be placed
 in that directory.
diff --git a/src/04-modality-specific-files/01-magnetic-resonance-imaging-data.md b/src/04-modality-specific-files/01-magnetic-resonance-imaging-data.md
index e09bb59eb6..823ec56f5f 100644
--- a/src/04-modality-specific-files/01-magnetic-resonance-imaging-data.md
+++ b/src/04-modality-specific-files/01-magnetic-resonance-imaging-data.md
@@ -167,7 +167,7 @@ whenever possible. See also
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -359,7 +359,7 @@ and a guide for using macros can be found at
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -567,7 +567,7 @@ and a guide for using macros can be found at
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -777,7 +777,7 @@ and can be used for practical guidance when curating a new dataset.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -1009,7 +1009,7 @@ containing that type of fieldmap can be found here:
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -1055,7 +1055,7 @@ second echos are available.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -1091,7 +1091,7 @@ In some cases (for example GE), the scanner software will directly reconstruct a
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -1140,7 +1140,7 @@ AFNI `3dqwarp`.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/04-modality-specific-files/02-magnetoencephalography.md b/src/04-modality-specific-files/02-magnetoencephalography.md
index 0368ba8fe2..6af408f776 100644
--- a/src/04-modality-specific-files/02-magnetoencephalography.md
+++ b/src/04-modality-specific-files/02-magnetoencephalography.md
@@ -18,7 +18,7 @@ the [BIDS examples repository](https://github.com/bids-standard/bids-examples).
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -203,7 +203,7 @@ Date time information MUST be expressed as indicated in [Units](../02-common-pri
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -287,7 +287,7 @@ UADC001 AUDIO V envelope of audio signal presented to participant
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -400,7 +400,7 @@ Photos of the anatomical landmarks and/or head localization coils
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -427,7 +427,7 @@ actual anatomical nasion: `sub-0001_ses-001_acq-NAS_photo.jpg`
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -510,7 +510,7 @@ A guide for using macros can be found at
 
 Example 2:
 
-One recording per session, stored within the session folder.
+One recording per session, stored within the session directory.
 
 <!-- This block generates a file tree.
 A guide for using macros can be found at
diff --git a/src/04-modality-specific-files/03-electroencephalography.md b/src/04-modality-specific-files/03-electroencephalography.md
index 7b5db5b4af..6aa12e1dac 100644
--- a/src/04-modality-specific-files/03-electroencephalography.md
+++ b/src/04-modality-specific-files/03-electroencephalography.md
@@ -14,7 +14,7 @@ and can be used for practical guidance when curating a new dataset.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -183,7 +183,7 @@ Date time information MUST be expressed as indicated in [Units](../02-common-pri
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -258,7 +258,7 @@ UADC001  MISC  n/a    envelope of audio signal        n/a           good    n/a
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -305,7 +305,7 @@ the recording.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -423,7 +423,7 @@ Photos of the anatomical landmarks and/or fiducials.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/04-modality-specific-files/04-intracranial-electroencephalography.md b/src/04-modality-specific-files/04-intracranial-electroencephalography.md
index 9e5de8b303..1bccd67678 100644
--- a/src/04-modality-specific-files/04-intracranial-electroencephalography.md
+++ b/src/04-modality-specific-files/04-intracranial-electroencephalography.md
@@ -14,7 +14,7 @@ and can be used for practical guidance when curating a new dataset.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -203,7 +203,7 @@ Date time information MUST be expressed as indicated in [Units](../02-common-pri
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -276,7 +276,7 @@ Example of free-form text for field `description`:
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -350,7 +350,7 @@ H01   27  -42  -21  5      AdTech
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -430,7 +430,7 @@ Example:
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/04-modality-specific-files/07-behavioral-experiments.md b/src/04-modality-specific-files/07-behavioral-experiments.md
index 4c27add5a4..50e025a0a8 100644
--- a/src/04-modality-specific-files/07-behavioral-experiments.md
+++ b/src/04-modality-specific-files/07-behavioral-experiments.md
@@ -2,7 +2,7 @@
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/04-modality-specific-files/09-positron-emission-tomography.md b/src/04-modality-specific-files/09-positron-emission-tomography.md
index 0c04f86612..5202926e94 100644
--- a/src/04-modality-specific-files/09-positron-emission-tomography.md
+++ b/src/04-modality-specific-files/09-positron-emission-tomography.md
@@ -62,7 +62,7 @@ In this example, tracer injection coincides with scan start.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -296,7 +296,7 @@ ses-02 59
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/04-modality-specific-files/10-microscopy.md b/src/04-modality-specific-files/10-microscopy.md
index f8e1ec771f..148b098850 100644
--- a/src/04-modality-specific-files/10-microscopy.md
+++ b/src/04-modality-specific-files/10-microscopy.md
@@ -20,7 +20,7 @@ Further Microscopy datasets are available:
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -433,7 +433,7 @@ in a sample.
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/appendices/file-collections.md b/src/appendices/file-collections.md
index 3a256b59af..56c675f2ec 100644
--- a/src/appendices/file-collections.md
+++ b/src/appendices/file-collections.md
@@ -24,7 +24,7 @@ included in this appendix:
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
@@ -59,7 +59,7 @@ and a guide for using macros can be found at
 
 <!--
 This block generates a filename templates.
-The inputs for this macro can be found in the folder
+The inputs for this macro can be found in the directory
   src/schema/rules/datatypes
 and a guide for using macros can be found at
  https://github.com/bids-standard/bids-specification/blob/master/macros_doc.md
diff --git a/src/schema/objects/extensions.yaml b/src/schema/objects/extensions.yaml
index 6bc111a99f..1a0dc4d299 100644
--- a/src/schema/objects/extensions.yaml
+++ b/src/schema/objects/extensions.yaml
@@ -52,7 +52,7 @@ chn:
     A file generated by KRISS MEG systems containing the position of the center of the MEG coils.
 
     Each experimental run on the KRISS system produces a file with extension `.kdf`.
-    Additional files that may be available in the same folder include
+    Additional files that may be available in the same directory include
     the digitized positions of the head points (`\_digitizer.txt`),
     the position of the center of the MEG coils (`.chn`),
     and the event markers (`.trg`).
@@ -70,9 +70,9 @@ dat:
     A fine-calibration file used for Neuromag/Elekta/MEGIN MEG recording hardware.
 CTF:
   value: .ds/
-  display_name: CTF MEG Dataset Folder
+  display_name: CTF MEG Dataset Directory
   description: |
-    A folder for MEG data, typically containing a `.meg4` file for the data and a `.res4` file for the resources.
+    A directory for MEG data, typically containing a `.meg4` file for the data and a `.res4` file for the resources.
 dlabelnii:
   value: .dlabel.nii
   display_name: CIFTI-2 Dense Label File
@@ -132,7 +132,7 @@ kdf:
     A KRISS (file with extension `.kdf`) file.
 
     Each experimental run on the KRISS system produces a file with extension `.kdf`.
-    Additional files that may be available in the same folder include
+    Additional files that may be available in the same directory include
     the digitized positions of the head points (`\_digitizer.txt`),
     the position of the center of the MEG coils (`.chn`),
     and the event markers (`.trg`).
@@ -152,7 +152,7 @@ mefd:
   value: .mefd/
   display_name: Multiscale Electrophysiology File Format Version 3.0
   description: |
-    A folder in the [MEF3](https://osf.io/e3sf9/) format.
+    A directory in the [MEF3](https://osf.io/e3sf9/) format.
 
     Each recording consists of a `.mefd` directory.
 mhd:
@@ -263,7 +263,7 @@ trg:
     A file generated by KRISS MEG systems containing the event markers.
 
     Each experimental run on the KRISS system produces a file with extension `.kdf`.
-    Additional files that may be available in the same folder include
+    Additional files that may be available in the same directory include
     the digitized positions of the head points (`\_digitizer.txt`),
     the position of the center of the MEG coils (`.chn`),
     and the event markers (`.trg`).
@@ -312,7 +312,7 @@ None:
     A file with no extension.
 Directory:
   value: /
-  display_name: Folder
+  display_name: Directory
   description: |
-    A folder with no extension.
+    A directory with no extension.
     Corresponds to BTi/4D data.
diff --git a/tools/schemacode/bidsschematools/render/text.py b/tools/schemacode/bidsschematools/render/text.py
index 9d7011c299..52787cb1c2 100644
--- a/tools/schemacode/bidsschematools/render/text.py
+++ b/tools/schemacode/bidsschematools/render/text.py
@@ -420,7 +420,7 @@ def append_filename_template_legend(text, pdf_format=False):
   """
 
     legend = f"""{info_str}
-- Filename entities or folders between square brackets
+- Filename entities or directories between square brackets
   (for example, `[_ses-<label>]`) are OPTIONAL.
 - Some entities may only allow specific values,
   in which case those values are listed in `<>`, separated by `|`.
diff --git a/tools/schemacode/bidsschematools/tests/test_rules.py b/tools/schemacode/bidsschematools/tests/test_rules.py
index 2446a0808a..c44634f814 100644
--- a/tools/schemacode/bidsschematools/tests/test_rules.py
+++ b/tools/schemacode/bidsschematools/tests/test_rules.py
@@ -68,7 +68,7 @@ def test_rule_objects(schema_obj):
 
             for i_use, use in enumerate(instance):
                 if use == "derivatives":
-                    # Skip derivatives folders, because the folder is treated as a "use" instead.
+                    # Skip derivatives dirs, because the dir is treated as a "use" instead.
                     continue
                 elif "[]" in use:
                     # Rules may reference metadata fields with lists.