diff --git a/.github/workflows/cldf-validation.yml b/.github/workflows/cldf-validation.yml index ef8eac5..b2f938e 100644 --- a/.github/workflows/cldf-validation.yml +++ b/.github/workflows/cldf-validation.yml @@ -12,7 +12,7 @@ jobs: runs-on: ubuntu-latest strategy: matrix: - python-version: [3.11] + python-version: [3.12] steps: - uses: actions/checkout@v2 diff --git a/.zenodo.json b/.zenodo.json index 318a08b..f982f47 100644 --- a/.zenodo.json +++ b/.zenodo.json @@ -18,6 +18,10 @@ "name": "Johann-Mattis List", "type": "Editor" }, + { + "name": "Christoph Rzymski", + "type": "Editor" + }, { "name": "Tiago Tresoldi", "type": "Other" diff --git a/CONTRIBUTORS.md b/CONTRIBUTORS.md index e83788a..07976ff 100644 --- a/CONTRIBUTORS.md +++ b/CONTRIBUTORS.md @@ -2,7 +2,8 @@ Name | GitHub user | Description | Role --- | --- | --- | --- -Johann-Mattis List | @lingulist | orthography profile | Editor +Johann-Mattis List | @lingulist | orthography profile | Editor +Christoph Rzymski | @chrzyki | patron | Editor Tiago Tresoldi | @tresoldi | code | Other Derek Nurse | | original data collection | Author Gerard Philippson | | original data collection | Author diff --git a/README.md b/README.md index 4429063..a01acbd 100644 --- a/README.md +++ b/README.md @@ -29,8 +29,8 @@ Conceptlists in Concepticon: ![BIPA: 100%](https://img.shields.io/badge/BIPA-100%25-brightgreen.svg "BIPA: 100%") ![CLTS SoundClass: 100%](https://img.shields.io/badge/CLTS%20SoundClass-100%25-brightgreen.svg "CLTS SoundClass: 100%") -- **Varieties:** 127 -- **Concepts:** 1,052 +- **Varieties:** 127 (linked to 107 different Glottocodes) +- **Concepts:** 1,052 (linked to 650 different Concepticon concept sets) - **Lexemes:** 115,642 - **Sources:** 4 - **Synonymy:** 1.01 @@ -43,7 +43,8 @@ Conceptlists in Concepticon: Name | GitHub user | Description | Role --- | --- | --- | --- -Johann-Mattis List | @lingulist | orthography profile | Editor +Johann-Mattis List | @lingulist | orthography profile | Editor +Christoph Rzymski | @chrzyki | patron | Editor Tiago Tresoldi | @tresoldi | code | Other Derek Nurse | | original data collection | Author Gerard Philippson | | original data collection | Author diff --git a/cldf/README.md b/cldf/README.md index 42e3fcb..f1b4668 100644 --- a/cldf/README.md +++ b/cldf/README.md @@ -14,8 +14,8 @@ property | value [dc:identifier](http://purl.org/dc/terms/identifier) | http://www.cbold.ish-lyon.cnrs.fr/Docs/TLSinfo.html [dc:license](http://purl.org/dc/terms/license) | https://creativecommons.org/licenses/by/4.0/ [dcat:accessURL](http://www.w3.org/ns/dcat#accessURL) | https://github.com/lexibank/tls -[prov:wasDerivedFrom](http://www.w3.org/ns/prov#wasDerivedFrom) |
  1. lexibank/tls v4.1.1
  2. Glottolog v4.8
  3. Concepticon v3.1.0
  4. CLTS v2.2.0
-[prov:wasGeneratedBy](http://www.w3.org/ns/prov#wasGeneratedBy) |
  1. lingpy-rcParams: lingpy-rcParams.json
  2. python: 3.10.12
  3. python-packages: requirements.txt
+[prov:wasDerivedFrom](http://www.w3.org/ns/prov#wasDerivedFrom) |
  1. lexibank/tls v4.1.1
  2. Glottolog v5.0
  3. Concepticon v3.2.0
  4. CLTS v2.3.0
+[prov:wasGeneratedBy](http://www.w3.org/ns/prov#wasGeneratedBy) |
  1. lingpy-rcParams: lingpy-rcParams.json
  2. python: 3.12.4
  3. python-packages: requirements.txt
[rdf:ID](http://www.w3.org/1999/02/22-rdf-syntax-ns#ID) | tls [rdf:type](http://www.w3.org/1999/02/22-rdf-syntax-ns#type) | http://www.w3.org/ns/dcat#Distribution diff --git a/cldf/cldf-metadata.json b/cldf/cldf-metadata.json index fead43b..82980e0 100644 --- a/cldf/cldf-metadata.json +++ b/cldf/cldf-metadata.json @@ -23,19 +23,19 @@ { "rdf:about": "https://github.com/glottolog/glottolog", "rdf:type": "prov:Entity", - "dc:created": "v4.8", + "dc:created": "v5.0", "dc:title": "Glottolog" }, { "rdf:about": "https://github.com/concepticon/concepticon-data", "rdf:type": "prov:Entity", - "dc:created": "v3.1.0", + "dc:created": "v3.2.0", "dc:title": "Concepticon" }, { "rdf:about": "https://github.com/cldf-clts/clts", "rdf:type": "prov:Entity", - "dc:created": "v2.2.0", + "dc:created": "v2.3.0", "dc:title": "CLTS" } ], @@ -46,7 +46,7 @@ }, { "dc:title": "python", - "dc:description": "3.10.12" + "dc:description": "3.12.4" }, { "dc:title": "python-packages", diff --git a/cldf/lingpy-rcParams.json b/cldf/lingpy-rcParams.json index e235795..f6a1b5e 100644 --- a/cldf/lingpy-rcParams.json +++ b/cldf/lingpy-rcParams.json @@ -64,7 +64,7 @@ 10, 10 ], - "filename": "lingpy-2023-12-21", + "filename": "lingpy-2024-08-01", "gap_symbol": "-", "gap_weight": 0.5, "gop": -2, @@ -123,7 +123,7 @@ "scorer": {}, "sonar": true, "stress": "\u02c8\u02cc'", - "timestamp": "2023-12-21 11:12", + "timestamp": "2024-08-01 11:06", "tones": "\u00b9\u00b2\u00b3\u2074\u2075\u2076\u2077\u2078\u2079\u2070\u2081\u2082\u2083\u2084\u2085\u2086\u2087\u2088\u2089\u20800123456789\u02e5\u02e6\u02e7\u02e8\u02e9\u02ea\u02eb-\ua708-\ua709-\ua70a-\ua70b-\ua70c-\ua70d-\ua70e-\ua70f-\ua710-\ua711-\ua712-\ua713-\ua714-\ua715-\ua716-\ua717-\ua718-\ua719-\ua71a-\ua700-\ua701-\ua702-\ua703-\ua704-\ua705-\ua706-\ua707", "tree_calc": "neighbor", "unique_sequences": true, diff --git a/cldf/parameters.csv b/cldf/parameters.csv index 1ae0079..65af380 100644 --- a/cldf/parameters.csv +++ b/cldf/parameters.csv @@ -323,7 +323,7 @@ ID,Name,Concepticon_ID,Concepticon_Gloss,Swahili_Gloss,NUMBER 316_garden,garden,586,GARDEN,bustani,316 317_hippopotamus,hippopotamus,,,kiboko,317 318_rhinoceros,rhinoceros,2591,RHINOCEROS,kifaru,318 -319_crossariver,cross (a river),1742,TRAVERSE,(ku)vuka,319 +319_crossariver,cross (a river),574,TRANSIT,(ku)vuka,319 320_hunger,hunger,1864,HUNGER,njaa,320 321_rubbishheap,rubbish heap,909,GARBAGE,jaa,321 322_water,water,948,WATER,maji,322 diff --git a/cldf/requirements.txt b/cldf/requirements.txt index b92a8a7..19ecbc1 100644 --- a/cldf/requirements.txt +++ b/cldf/requirements.txt @@ -1,59 +1,56 @@ appdirs==1.4.4 -attrs==23.1.0 -Babel==2.14.0 -bibtexparser==2.0.0b4 -bs4==0.0.1 -certifi==2023.11.17 +attrs==23.2.0 +Babel==2.15.0 +bibtexparser==2.0.0b7 +bs4==0.0.2 +certifi==2024.7.4 cldfbench==1.14.0 cldfcatalog==1.5.1 -cldfzenodo==2.1.0 -clldutils==3.22.1 +cldfzenodo==2.1.1 +clldutils==3.22.2 colorama==0.4.6 -colorlog==6.8.0 -csvw==3.2.1 -exceptiongroup==1.2.0 +colorlog==6.8.2 +csvw==3.3.0 gitdb==4.0.11 -greenlet==3.0.2 -idna==3.6 -idspy==0.3.0 +greenlet==3.0.3 +idna==3.7 iniconfig==2.0.0 isodate==0.6.1 -jsonschema==4.20.0 +jsonschema==4.23.0 lingpy==2.6.13 -lxml==4.9.3 -Markdown==3.5.1 +lxml==5.2.2 +Markdown==3.6 nameparser==1.1.3 -networkx==3.2.1 +networkx==3.3 newick==1.9.0 -numpy==1.26.2 -openpyxl==3.1.2 -packaging==23.2 -pluggy==1.3.0 -purl==1.6 +numpy==2.0.1 +openpyxl==3.1.5 +packaging==24.1 +pluggy==1.5.0 pybtex==0.24.0 -pycldf==1.36.0 -pyclts==3.1.1 -pyconcepticon==3.0.0 -pycountry==23.12.11 -pyglottolog==3.12.0 +pycldf==1.38.1 +pyclts==3.2.0 +pyconcepticon==3.1.0 +pycountry==24.6.1 +pyglottolog==3.13.0 pylatexenc==2.10 -pylexibank==3.4.0 -pytest==7.4.3 -python-dateutil==2.8.2 +pylexibank==3.5.0 +pytest==8.3.2 +python-dateutil==2.9.0.post0 rdflib==7.0.0 -referencing==0.32.0 -regex==2023.10.3 -requests==2.31.0 +referencing==0.35.1 +regex==2024.7.24 +requests==2.32.3 rfc3986==1.5.0 segments==2.2.1 six==1.16.0 smmap==5.0.1 soupsieve==2.5 -SQLAlchemy==1.4.50 +SQLAlchemy==1.4.53 tabulate==0.9.0 termcolor==2.4.0 -tqdm==4.66.1 +tqdm==4.66.4 uritemplate==4.1.1 -urllib3==2.1.0 +urllib3==2.2.2 xlrd==2.0.1 zenodoclient==0.5.1 \ No newline at end of file diff --git a/lexibank_tls.py b/lexibank_tls.py index 57492b5..110920c 100644 --- a/lexibank_tls.py +++ b/lexibank_tls.py @@ -18,6 +18,8 @@ class CustomConcept(pylexibank.Concept): class Dataset(pylexibank.Dataset): dir = Path(__file__).parent id = "tls" + writer_options = dict(keep_languages=False, keep_parameters=False) + concept_class = CustomConcept form_spec = pylexibank.FormSpec( brackets={"(": ")", "[": "]"}, @@ -25,7 +27,9 @@ class Dataset(pylexibank.Dataset): missing_data=("-", "?", "???", "+", "_", "--_", "!"), replacements=[ ("(kU)d\x97s\x87 ? +", ""), - ("\x88", ""), ("\x87", ""), ("\x97", ""), + ("\x88", ""), + ("\x87", ""), + ("\x97", ""), (" ", "_"), ("#__#_pencil_only_#", ""), ('"', ""), @@ -45,9 +49,8 @@ class Dataset(pylexibank.Dataset): ("right_person", "-"), ("shitere_shi_shilenj..", "shitere"), ("#", ""), - - ], - first_form_only=True + ], + first_form_only=True, ) def cmd_makecldf(self, args):