brennercruvinel commited on
Commit
d0ecaa1
·
unverified ·
1 Parent(s): 2dec7ac

Regenerate the card in the conventional dataset card order

Browse files

Generated with build_card.py from 3char-pipeline; card_check.py passes. No Parquet file changed.

Files changed (2) hide show
  1. CHANGELOG.md +1 -1
  2. README.md +114 -121
CHANGELOG.md CHANGED
@@ -6,7 +6,7 @@ Documentation only. No Parquet file has changed since 0.1.1.
6
 
7
  ### Card
8
 
9
- The card was restructured as the description of a scientific data resource. It now opens with a definition of the collection and the reason it exists, describes the five kinds of list it contains (assigned codes, symbols, vocabulary terms, community conventions and test vectors), and then covers uses, loading, construction, findings, licensing and limitations. The full inventory of the 58 systems moved to an appendix that lists each system's kind, authority and size; regular expressions and the other metadata remain in the `systems` configuration. System names are displayed with an initial capital, while the stored values are unchanged. The citation title now reads "58 code systems and controlled vocabularies" instead of "58 identifier systems".
10
 
11
  A new section, with a figure, shows which lists describe the same entities, under the same codes or under different ones, and how many shared codes carry the same name in two systems.
12
 
 
6
 
7
  ### Card
8
 
9
+ The card was restructured as the description of a scientific data resource. It now opens with a definition of the collection and the reason it exists, describes the five kinds of list it contains (assigned codes, symbols, vocabulary terms, community conventions and test vectors), and then covers uses, loading, construction, findings, licensing and limitations. The card follows the usual order of a dataset card (contents, uses, structure, construction, analyses, limitations, licensing, related resources, citation) and ends with the full inventory of the 58 systems, listing each system's kind, authority, size and regular expression. System names are displayed with an initial capital, while the stored values are unchanged. The citation title now reads "58 code systems and controlled vocabularies" instead of "58 identifier systems".
10
 
11
  A new section, with a figure, shows which lists describe the same entities, under the same codes or under different ones, and how many shared codes carry the same name in two systems.
12
 
README.md CHANGED
@@ -258,13 +258,11 @@ configs:
258
  ---
259
  # 3char
260
 
261
- 3char is a curated collection of code systems, identifiers, symbols and controlled vocabularies from documented reference sources. It brings 751,870 heterogeneous records from 58 sources into a common Parquet schema, while preserving each record's source namespace and, for every source, its provenance, format constraints and licensing information.
262
 
263
- Short codes are reused across sources. `BRA` is Brazil in both ISO 3166-1 and the IOC country list, the Braj language in the IANA Language Subtag Registry, and an airport in the IATA list. Registries of identifier systems such as the Bioregistry describe these systems without redistributing their contents, and Wikidata records identifiers as statements about entities rather than as the lists their authorities publish. Keeping the lists side by side, each in its own namespace, makes it possible to see where the same string occurs in different sources, which lists describe the same entities with the same or with different codes, how each code compares with the form its source defines, and under which terms each list may be reused.
264
 
265
- This card describes release 0.1.3. The data files are unchanged since release 0.1.1 (2026-09-12).
266
-
267
- ## What the collection contains
268
 
269
  The 58 sources span ten domains, from biology and finance to language, transport and time. They are not all the same kind of list, and the difference matters when a record is read:
270
 
@@ -276,7 +274,7 @@ The 58 sources span ten domains, from biology and finance to language, transport
276
  | Conventions maintained by a community | 2 | File extensions, FourCC codec identifiers |
277
  | Test vectors | 1 | The reference test cases of the Open Location Code encoder |
278
 
279
- Entries also differ in status within a list. The HTTP status code registry is included together with its unassigned values, one row per unassigned code (436 rows, marked `unassigned`), and five sources keep entries they mark as no longer in use (6,793 rows, marked `deprecated`). The [appendix](#appendix-all-58-systems) lists every system with its kind, authority and size. The `systems` configuration holds the full metadata of each system, including its regular expression, license, version and retrieval date.
280
 
281
  The name comes from the three-character codes, such as airport, currency and country codes, with which the collection began. Codes in the collection range from one character to media types of 84.
282
 
@@ -284,12 +282,12 @@ The name comes from the three-character codes, such as airport, currency and cou
284
 
285
  - **Validating and normalizing codes.** Each list and its regular expression show whether a string is an entry in the recorded edition of the list, and whether it has the form the source defines.
286
  - **Grounding language models and search.** The lists can serve as lookup tables or retrieval corpora that map codes to names and back. [Retrieval](#retrieval) summarizes how the presentation of records affected retrieval in a controlled sample.
287
- - **Studying shared codes and equivalent lists.** The collision census identifies every string that occurs in more than one system, and [Lists that describe the same entities](#lists-that-describe-the-same-entities) shows which lists can be mapped onto each other. Whether a shared string denotes the same entity has to be established for each pair of systems: `BRA` names Brazil in two of them and unrelated entities in others. The closed study [poc-nsi](https://github.com/brennercruvinel/poc-nsi) used seven systems in this way to test whether language models confuse such codes.
288
  - **Linking tables.** Country codes stored in `extra`, such as `extra.country` in the airport systems and `extra.iso3` in `lng/ioc`, connect several systems to ISO 3166-1.
289
 
290
  For operational work that depends on the latest edition of a list, such as billing with ICD-10-CM, compare `version` and `retrieved_at` with the source's current release.
291
 
292
- ## Using the data
293
 
294
  ```python
295
  from datasets import load_dataset
@@ -338,7 +336,7 @@ Every system has the same Arrow schema, with one row per entry:
338
 
339
  Across systems, a record is identified by `(system_id, code)`, or by `(system_id, extra.locode)` in `trp/unl`. A code without its system does not identify a record.
340
 
341
- ## How the collection was built
342
 
343
  The data is produced by [3char-pipeline](https://github.com/brennercruvinel/3char-pipeline), which downloads each source, parses it into the common schema and checks the result. A source is included when four conditions hold, each checked by a column or a file in the pipeline:
344
 
@@ -357,9 +355,9 @@ Eight systems first taken from secondary copies (gists, unlicensed repositories,
357
 
358
  Every build verifies each raw file against its recorded checksum and tests each entry against its system's regular expression; either check stops the build on a mismatch. Only systems whose `license_status` is `ok` or `check` are published (see [Licensing](#licensing)).
359
 
360
- ## What the collection shows
361
 
362
- Every figure in this section is produced by a script in the pipeline (`census.py`, `keys.py`, `equivalences.py`, `crosscheck_wikidata.py`) and checked against this card by `card_check.py`.
363
 
364
  ### Codes shared between systems
365
 
@@ -373,7 +371,7 @@ Codes are compared after case folding, and a code is shared when the same folded
373
  | In eleven systems, the maximum | 2 (`and`, `car`) |
374
  | In two or more systems, counting the derived pair `lng/b47` and `lng/i63` once | 14,845 |
375
 
376
- A shared string is not necessarily a shared meaning. Some systems describe the same entities with the same codes, as ISO 3166-1 alpha-3 and the IOC list do for many countries, and `lng/i63` is derived from `lng/b47`. Others share strings by coincidence: `Arg` is arginine among the amino acid codes, while `ARG` is Argentina in the country lists and an airport in the IATA list. The census counts shared strings; it does not decide whether they refer to the same entity.
377
 
378
  Every code in eight systems also occurs in at least one other system: `bio/aa1`, `bio/aa3`, `bio/cdn`, `lng/i31`, `lng/i39`, `lng/i63`, `lng/mdy` and `wrt/iau`. In `trp/iata`, `lng/ioc`, `lng/i15`, `trp/ica2` and `fin/i42`, more than 98% of codes do. Without case folding there are 632,457 distinct strings, of which 21,255 occur in more than one system. Within `trp/unl`, 18,748 codes repeat because the key is the country plus the code; these repetitions are not counted as sharing between systems.
379
 
@@ -426,30 +424,7 @@ In each divergence, the rejected codes appear in the source's current file and t
426
 
427
  In a separate test, a bare code shared by several sampled systems retrieved a record of the intended system in 33.0% of 182 cases, the rate of a random choice among those systems; with the system name in both the chunk and the query, it did so in all cases. With 51 questions, one question is worth 1.96 points of recall, and the results describe this sample only.
428
 
429
- ## Licensing
430
-
431
- The dataset has no single license. Each row of the `systems` configuration records `license` and `license_status`. The `license` column takes 27 distinct values: an SPDX identifier where one applies, and otherwise a short tag defined in `LICENSES.md`. Of the 58 published systems, 56 have the status `ok` and 2 have the status `check`.
432
-
433
- - **No stated terms (`check`).** `fin/cfi` and `fin/i42` come from SIX Group, which publishes both lists free of charge but states no terms for redistribution, so their redistribution status is unconfirmed. They are included because the codes are plain facts in worldwide use; this is the curator's judgment, not a grant of rights. Anyone who redistributes them commercially should first obtain SIX Group's terms.
434
- - **Copyleft and share-alike (seven systems).** `cmp/frc` (GPL-2.0-or-later) and `cmp/git` (GPL-2.0-only) are lists extracted from GPL source trees; `cmp/vim` is under the Vim license; `grf/sch` and `grf/sio` are under CC-BY-SA-3.0; `lng/i31` is under CC-BY-SA-4.0, the license of its mirror; and `trp/ica2` is under ODbL-1.0. Their attribution requirements apply to any redistribution, and the share-alike terms also apply to redistributed copies of the tables.
435
- - **ISO code lists.** `lng/i31` and `lng/i39` have the status `ok` on the basis of their mirrors' licenses. ISO asserts copyright over its code lists, and a mirror's license does not change the upstream terms.
436
- - **Excluded systems.** The WHO ATC index, the BISAC subject headings, the Dewey summaries, the CUSIP mapping, the ISO 10383 MIC list and the what3words API description are parsed but never published, because their terms do not permit redistribution. The SIL ISO 639-3 table was removed for the same reason; `lng/i63` is a different list, the three-letter subtags of the IANA registry.
437
-
438
- Only Parquet files are published. Raw source files are downloaded by the pipeline and are not stored in either repository.
439
-
440
- ## Relation to existing resources
441
-
442
- | Resource | Contents | Relation to 3char |
443
- | --- | --- | --- |
444
- | [Bioregistry](https://bioregistry.io) (2,936 prefixes, snapshot of 2026-10-13) | A registry of identifier systems, mostly biomedical, with prefixes, regular expressions and resolvers | Describes systems without redistributing their contents; 3char distributes the lists, across ten domains |
445
- | [N2T](https://n2t.net) (511 prefix entries in the full list) | Resolver prefixes | As for the Bioregistry. Four 3char IDs (`elm`, `icd`, `mim`, `pdb`) are also N2T prefixes, and two of them denote other systems there |
446
- | Wikidata external-identifier properties (9,846 with a format constraint, 2026-10-13) | One property per system, with a regular expression and identifiers attached to entities | Identifiers are statements about entities rather than the authority's list; the two are compared in [Agreement with Wikidata format constraints](#agreement-with-wikidata-format-constraints) |
447
- | pycountry 26.2.16 | ISO 3166-1 (249), ISO 3166-2 (5,046), ISO 4217 (178), ISO 639 (7,923) and ISO 15924 (226), as a Python package built from the Debian iso-codes files | Covers five ISO standards; 3char takes ISO 3166-1, 4217, 639-1 and 15924 from other sources and includes 54 other systems |
448
- | `datasets/country-codes`, OurAirports, OpenFlights, IANA registries | Source files for single domains | Sources of individual 3char systems, cited in `systems.url` |
449
-
450
- A search of the Hugging Face Hub on 2026-10-14, with queries that included ISO 3166, ISO 4217, IATA, ICAO, ICD-10, HGNC, LOCODE, country codes, currency codes, identifiers, code systems, controlled vocabulary and three letter, returned 29 datasets, each covering a single system. No multi-domain collection of code lists with per-system license status was found. The search was not exhaustive, and this is not a claim of priority.
451
-
452
- ## Known limitations
453
 
454
  Labels come from each source and are not reconciled across systems: the same airport is `São Paulo/Guarulhos–Governor André Franco Montoro International Airport` in `trp/iata` and `trp/icao`, and `Guarulhos Apt / São Paulo`, under `BRGRU`, in `trp/unl`. Airports have no city-name alias.
455
 
@@ -473,6 +448,29 @@ Labels come from each source and are not reconciled across systems: the same air
473
 
474
  Coverage follows what sources publish under terms that permit redistribution. Standards bodies that publish in English and computing registries are well represented, while national schemes outside the United States and the UN system are not yet included. The collision census describes this collection, so adding a system changes it.
475
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
476
  ## Versions and citation
477
 
478
  Releases are numbered `MAJOR.MINOR.PATCH`. A change to the schema increments the major version, a change to any Parquet file increments the minor version, and a change limited to the card or the changelog increments the patch version. The changelog is maintained in the pipeline repository and copied here.
@@ -489,119 +487,114 @@ Release tags on the Hub and a Zenodo deposit are planned but have not been creat
489
  }
490
  ```
491
 
492
- ## Appendix: all 58 systems
493
 
494
- Kinds are those described in [What the collection contains](#what-the-collection-contains). Regular expressions, licenses, versions, retrieval dates and curator notes are in the `systems` configuration.
495
-
496
- <details>
497
- <summary>Systems by domain</summary>
498
 
499
  ### Biology (`bio`)
500
 
501
- | ID | System | Kind | Authority | Codes |
502
- | --- | --- | --- | --- | ---: |
503
- | `bio/aa1` | Amino acid 1-letter codes | Symbol | IUPAC-IUB JCBN | 20 |
504
- | `bio/aa3` | Amino acid 3-letter codes | Symbol | IUPAC-IUB JCBN | 20 |
505
- | `bio/cdn` | NCBI genetic codes (codons) | Symbol | NCBI | 64 |
506
- | `bio/ec` | Enzyme commission numbers | Assigned code | IUBMB / ExPASy | 8,441 |
507
- | `bio/gen` | HGNC gene symbols | Assigned code | HUGO Gene Nomenclature Committee | 44,997 |
508
- | `bio/pdb` | PDB entry IDs | Assigned code | wwPDB / RCSB | 254,978 |
509
 
510
  ### Classification (`cls`)
511
 
512
- | ID | System | Kind | Authority | Codes |
513
- | --- | --- | --- | --- | ---: |
514
- | `cls/elm` | Chemical element symbols | Symbol | IUPAC | 118 |
515
- | `cls/icd` | ICD-10-CM | Assigned code | US CDC / NCHS | 98,403 |
516
- | `cls/loc` | Library of Congress classification outline | Assigned code | Library of Congress | 234 |
517
 
518
  ### Computing (`cmp`)
519
 
520
- | ID | System | Kind | Authority | Codes |
521
- | --- | --- | --- | --- | ---: |
522
- | `cmp/asm` | x86 instruction mnemonics | Vocabulary term | Intel / AMD | 1,309 |
523
- | `cmp/css` | CSS functions | Vocabulary term | W3C CSSWG / MDN | 105 |
524
- | `cmp/dns` | DNS resource record types | Assigned code | IANA | 97 |
525
- | `cmp/ern` | POSIX errno names | Assigned code | IEEE / The Open Group | 77 |
526
- | `cmp/ext` | File extensions | Community convention | Community | 286 |
527
- | `cmp/frc` | FourCC codec identifiers | Community convention | Community / VLC | 672 |
528
- | `cmp/git` | Git subcommands | Vocabulary term | Git project | 158 |
529
- | `cmp/hex` | CSS named colors | Vocabulary term | W3C CSSWG | 148 |
530
- | `cmp/htt` | HTTP methods | Assigned code | IANA | 41 |
531
- | `cmp/mim` | Media types | Assigned code | IANA | 2,300 |
532
- | `cmp/reg` | x86-64 register names | Vocabulary term | LLVM Project | 257 |
533
- | `cmp/sts` | HTTP status codes | Assigned code | IANA | 500 |
534
- | `cmp/vim` | Vim motion commands | Vocabulary term | Vim project | 289 |
535
 
536
  ### Communication (`com`)
537
 
538
- | ID | System | Kind | Authority | Codes |
539
- | --- | --- | --- | --- | ---: |
540
- | `com/bau` | Baudot / ITA2 teleprinter codes | Symbol | ITU-T | 30 |
541
- | `com/mor` | Morse code | Symbol | ITU-R | 48 |
542
- | `com/tlg` | ABC universal commercial telegraphic code | Symbol | William Clauson-Thue (1901) | 7,359 |
543
 
544
  ### Finance (`fin`)
545
 
546
- | ID | System | Kind | Authority | Codes |
547
- | --- | --- | --- | --- | ---: |
548
- | `fin/cfi` | ISO 10962 CFI codes | Assigned code | SIX Group / ISO | 1,216 |
549
- | `fin/i42` | ISO 4217 currency codes | Assigned code | SIX Group / ISO | 178 |
550
- | `fin/tck` | SEC company tickers | Assigned code | US SEC | 10,405 |
551
 
552
  ### Graph and vocabulary (`grf`)
553
 
554
- | ID | System | Kind | Authority | Codes |
555
- | --- | --- | --- | --- | ---: |
556
- | `grf/act` | ActivityStreams 2.0 terms | Vocabulary term | W3C | 141 |
557
- | `grf/foa` | FOAF vocabulary terms | Vocabulary term | FOAF project | 75 |
558
- | `grf/gql` | openCypher keywords | Vocabulary term | openCypher / Neo4j | 63 |
559
- | `grf/lpg` | Property graph schema terms | Vocabulary term | arXiv 2211.10962 | 34 |
560
- | `grf/oid` | OpenID Connect discovery fields | Vocabulary term | OpenID Foundation | 17 |
561
- | `grf/org` | W3C Organization ontology terms | Vocabulary term | W3C | 45 |
562
- | `grf/rdf` | RDF Schema terms | Vocabulary term | W3C | 15 |
563
- | `grf/sch` | schema.org types and properties | Vocabulary term | schema.org / W3C CG | 2,987 |
564
- | `grf/shc` | SHACL vocabulary terms | Vocabulary term | W3C | 221 |
565
- | `grf/sio` | SIOC vocabulary terms | Vocabulary term | SIOC Project / DERI | 99 |
566
- | `grf/vcd` | vCard ontology terms | Vocabulary term | W3C | 146 |
567
 
568
  ### Language (`lng`)
569
 
570
- | ID | System | Kind | Authority | Codes |
571
- | --- | --- | --- | --- | ---: |
572
- | `lng/b47` | BCP 47 language subtags | Assigned code | IANA | 9,295 |
573
- | `lng/glt` | Glottolog languoid codes | Assigned code | MPI-EVA Leipzig | 27,177 |
574
- | `lng/i15` | ISO 15924 script codes | Assigned code | Unicode Consortium (RA) | 226 |
575
- | `lng/i31` | ISO 3166-1 country codes | Assigned code | ISO 3166/MA | 249 |
576
- | `lng/i39` | ISO 639-1 language codes | Assigned code | ISO 639/RA | 183 |
577
- | `lng/i63` | BCP 47 three-letter language subtags | Assigned code | IANA | 8,084 |
578
- | `lng/ioc` | Country codes composite | Assigned code | Community / datasets.io | 226 |
579
- | `lng/mdy` | Month names | Vocabulary term | Unicode CLDR | 12 |
580
 
581
  ### Lexical (`nte`)
582
 
583
- | ID | System | Kind | Authority | Codes |
584
- | --- | --- | --- | --- | ---: |
585
- | `nte/wnt` | English WordNet 2024 synset IDs | Assigned code | Global WordNet Association | 120,630 |
586
 
587
  ### Transport and place (`trp`)
588
 
589
- | ID | System | Kind | Authority | Codes |
590
- | --- | --- | --- | --- | ---: |
591
- | `trp/iata` | IATA airport codes | Assigned code | IATA | 9,056 |
592
- | `trp/ica2` | IATA and ICAO airline codes | Assigned code | IATA / ICAO | 5,841 |
593
- | `trp/icao` | ICAO airport codes | Assigned code | ICAO | 10,153 |
594
- | `trp/olc` | Open Location Code (plus codes) | Test vector | Google | 292 |
595
- | `trp/tld` | Top level domains | Assigned code | IANA | 1,437 |
596
- | `trp/unl` | UN/LOCODE location codes | Assigned code | UNECE | 116,213 |
597
 
598
  ### Writing and time (`wrt`)
599
 
600
- | ID | System | Kind | Authority | Codes |
601
- | --- | --- | --- | --- | ---: |
602
- | `wrt/abj` | Hebrew script characters | Symbol | Unicode / community | 84 |
603
- | `wrt/iau` | IAU constellation abbreviations | Assigned code | International Astronomical Union | 88 |
604
- | `wrt/tz` | tz database zone names and abbreviations | Assigned code | IANA / tzdb | 806 |
605
- | `wrt/uni` | Unicode emoji sequences | Symbol | Unicode Consortium | 5,225 |
606
-
607
- </details>
 
258
  ---
259
  # 3char
260
 
261
+ 3char is a curated collection of code systems, identifiers, symbols and controlled vocabularies from documented reference sources. It gathers 751,870 records from 58 sources into one Parquet schema: ISO country, currency and language codes, IATA and ICAO airport codes, UN/LOCODE, ICD-10-CM, HGNC gene symbols, chemical element symbols, IANA registries, W3C vocabularies and others. Every record keeps the namespace of its source, and every source is documented with its provenance, version, format constraint and license.
262
 
263
+ The collection is meant for work that needs these lists as data rather than as documentation: validating and normalizing codes, grounding language models and search systems in authoritative lists, and studying how codes and lists relate across sources. This card describes release 0.1.3; the data files are unchanged since release 0.1.1 (2026-09-12).
264
 
265
+ ## Contents
 
 
266
 
267
  The 58 sources span ten domains, from biology and finance to language, transport and time. They are not all the same kind of list, and the difference matters when a record is read:
268
 
 
274
  | Conventions maintained by a community | 2 | File extensions, FourCC codec identifiers |
275
  | Test vectors | 1 | The reference test cases of the Open Location Code encoder |
276
 
277
+ Entries also differ in status within a list. The HTTP status code registry is included together with its unassigned values, one row per unassigned code (436 rows, marked `unassigned`), and five sources keep entries they mark as no longer in use (6,793 rows, marked `deprecated`). Every system is listed, with its kind, authority, size and regular expression, in [All 58 systems](#all-58-systems) at the end of this card; the `systems` configuration holds the rest of its metadata.
278
 
279
  The name comes from the three-character codes, such as airport, currency and country codes, with which the collection began. Codes in the collection range from one character to media types of 84.
280
 
 
282
 
283
  - **Validating and normalizing codes.** Each list and its regular expression show whether a string is an entry in the recorded edition of the list, and whether it has the form the source defines.
284
  - **Grounding language models and search.** The lists can serve as lookup tables or retrieval corpora that map codes to names and back. [Retrieval](#retrieval) summarizes how the presentation of records affected retrieval in a controlled sample.
285
+ - **Studying codes across lists.** The same string often occurs in several lists, sometimes for the same entity and sometimes not; [Analyses](#analyses) measures both. The closed study [poc-nsi](https://github.com/brennercruvinel/poc-nsi) used seven systems in this way to test whether language models confuse such codes.
286
  - **Linking tables.** Country codes stored in `extra`, such as `extra.country` in the airport systems and `extra.iso3` in `lng/ioc`, connect several systems to ISO 3166-1.
287
 
288
  For operational work that depends on the latest edition of a list, such as billing with ICD-10-CM, compare `version` and `retrieved_at` with the source's current release.
289
 
290
+ ## Dataset structure
291
 
292
  ```python
293
  from datasets import load_dataset
 
336
 
337
  Across systems, a record is identified by `(system_id, code)`, or by `(system_id, extra.locode)` in `trp/unl`. A code without its system does not identify a record.
338
 
339
+ ## Construction
340
 
341
  The data is produced by [3char-pipeline](https://github.com/brennercruvinel/3char-pipeline), which downloads each source, parses it into the common schema and checks the result. A source is included when four conditions hold, each checked by a column or a file in the pipeline:
342
 
 
355
 
356
  Every build verifies each raw file against its recorded checksum and tests each entry against its system's regular expression; either check stops the build on a mismatch. Only systems whose `license_status` is `ok` or `check` are published (see [Licensing](#licensing)).
357
 
358
+ ## Analyses
359
 
360
+ These analyses use the published files. Every figure in this section is produced by a script in the pipeline (`census.py`, `keys.py`, `equivalences.py`, `crosscheck_wikidata.py`) and checked against this card by `card_check.py`.
361
 
362
  ### Codes shared between systems
363
 
 
371
  | In eleven systems, the maximum | 2 (`and`, `car`) |
372
  | In two or more systems, counting the derived pair `lng/b47` and `lng/i63` once | 14,845 |
373
 
374
+ A shared string is not necessarily a shared meaning. `BRA` is Brazil in both ISO 3166-1 and the IOC list, the Braj language in the IANA Language Subtag Registry, and an airport in the IATA list. The census counts shared strings; whether they refer to the same entity is examined in the next section.
375
 
376
  Every code in eight systems also occurs in at least one other system: `bio/aa1`, `bio/aa3`, `bio/cdn`, `lng/i31`, `lng/i39`, `lng/i63`, `lng/mdy` and `wrt/iau`. In `trp/iata`, `lng/ioc`, `lng/i15`, `trp/ica2` and `fin/i42`, more than 98% of codes do. Without case folding there are 632,457 distinct strings, of which 21,255 occur in more than one system. Within `trp/unl`, 18,748 codes repeat because the key is the country plus the code; these repetitions are not counted as sharing between systems.
377
 
 
424
 
425
  In a separate test, a bare code shared by several sampled systems retrieved a record of the intended system in 33.0% of 182 cases, the rate of a random choice among those systems; with the system name in both the chunk and the query, it did so in all cases. With 51 questions, one question is worth 1.96 points of recall, and the results describe this sample only.
426
 
427
+ ## Limitations
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
428
 
429
  Labels come from each source and are not reconciled across systems: the same airport is `São Paulo/Guarulhos–Governor André Franco Montoro International Airport` in `trp/iata` and `trp/icao`, and `Guarulhos Apt / São Paulo`, under `BRGRU`, in `trp/unl`. Airports have no city-name alias.
430
 
 
448
 
449
  Coverage follows what sources publish under terms that permit redistribution. Standards bodies that publish in English and computing registries are well represented, while national schemes outside the United States and the UN system are not yet included. The collision census describes this collection, so adding a system changes it.
450
 
451
+ ## Licensing
452
+
453
+ The dataset has no single license. Each row of the `systems` configuration records `license` and `license_status`. The `license` column takes 27 distinct values: an SPDX identifier where one applies, and otherwise a short tag defined in `LICENSES.md`. Of the 58 published systems, 56 have the status `ok` and 2 have the status `check`.
454
+
455
+ - **No stated terms (`check`).** `fin/cfi` and `fin/i42` come from SIX Group, which publishes both lists free of charge but states no terms for redistribution, so their redistribution status is unconfirmed. They are included because the codes are plain facts in worldwide use; this is the curator's judgment, not a grant of rights. Anyone who redistributes them commercially should first obtain SIX Group's terms.
456
+ - **Copyleft and share-alike (seven systems).** `cmp/frc` (GPL-2.0-or-later) and `cmp/git` (GPL-2.0-only) are lists extracted from GPL source trees; `cmp/vim` is under the Vim license; `grf/sch` and `grf/sio` are under CC-BY-SA-3.0; `lng/i31` is under CC-BY-SA-4.0, the license of its mirror; and `trp/ica2` is under ODbL-1.0. Their attribution requirements apply to any redistribution, and the share-alike terms also apply to redistributed copies of the tables.
457
+ - **ISO code lists.** `lng/i31` and `lng/i39` have the status `ok` on the basis of their mirrors' licenses. ISO asserts copyright over its code lists, and a mirror's license does not change the upstream terms.
458
+ - **Excluded systems.** The WHO ATC index, the BISAC subject headings, the Dewey summaries, the CUSIP mapping, the ISO 10383 MIC list and the what3words API description are parsed but never published, because their terms do not permit redistribution. The SIL ISO 639-3 table was removed for the same reason; `lng/i63` is a different list, the three-letter subtags of the IANA registry.
459
+
460
+ Only Parquet files are published. Raw source files are downloaded by the pipeline and are not stored in either repository.
461
+
462
+ ## Related resources
463
+
464
+ | Resource | Contents | Relation to 3char |
465
+ | --- | --- | --- |
466
+ | [Bioregistry](https://bioregistry.io) (2,936 prefixes, snapshot of 2026-10-13) | A registry of identifier systems, mostly biomedical, with prefixes, regular expressions and resolvers | Describes systems without redistributing their contents; 3char distributes the lists, across ten domains |
467
+ | [N2T](https://n2t.net) (511 prefix entries in the full list) | Resolver prefixes | As for the Bioregistry. Four 3char IDs (`elm`, `icd`, `mim`, `pdb`) are also N2T prefixes, and two of them denote other systems there |
468
+ | Wikidata external-identifier properties (9,846 with a format constraint, 2026-10-13) | One property per system, with a regular expression and identifiers attached to entities | Identifiers are statements about entities rather than the authority's list; the two are compared in [Agreement with Wikidata format constraints](#agreement-with-wikidata-format-constraints) |
469
+ | pycountry 26.2.16 | ISO 3166-1 (249), ISO 3166-2 (5,046), ISO 4217 (178), ISO 639 (7,923) and ISO 15924 (226), as a Python package built from the Debian iso-codes files | Covers five ISO standards; 3char takes ISO 3166-1, 4217, 639-1 and 15924 from other sources and includes 54 other systems |
470
+ | `datasets/country-codes`, OurAirports, OpenFlights, IANA registries | Source files for single domains | Sources of individual 3char systems, cited in `systems.url` |
471
+
472
+ A search of the Hugging Face Hub on 2026-10-14, with queries that included ISO 3166, ISO 4217, IATA, ICAO, ICD-10, HGNC, LOCODE, country codes, currency codes, identifiers, code systems, controlled vocabulary and three letter, returned 29 datasets, each covering a single system. No multi-domain collection of code lists with per-system license status was found. The search was not exhaustive, and this is not a claim of priority.
473
+
474
  ## Versions and citation
475
 
476
  Releases are numbered `MAJOR.MINOR.PATCH`. A change to the schema increments the major version, a change to any Parquet file increments the minor version, and a change limited to the card or the changelog increments the patch version. The changelog is maintained in the pipeline repository and copied here.
 
487
  }
488
  ```
489
 
490
+ ## All 58 systems
491
 
492
+ Kinds are those described in [Contents](#contents). Licenses, versions, retrieval dates and curator notes are in the `systems` configuration.
 
 
 
493
 
494
  ### Biology (`bio`)
495
 
496
+ | ID | System | Kind | Authority | Codes | Pattern |
497
+ | --- | --- | --- | --- | ---: | --- |
498
+ | `bio/aa1` | Amino acid 1-letter codes | Symbol | IUPAC-IUB JCBN | 20 | `^[A-Z]$` |
499
+ | `bio/aa3` | Amino acid 3-letter codes | Symbol | IUPAC-IUB JCBN | 20 | `^[A-Z][a-z]{2}$` |
500
+ | `bio/cdn` | NCBI genetic codes (codons) | Symbol | NCBI | 64 | `^[ACGTU]{3}$` |
501
+ | `bio/ec` | Enzyme commission numbers | Assigned code | IUBMB / ExPASy | 8,441 | `^\d+\.\d+\.\d+\.(n?\d+\|-)$` |
502
+ | `bio/gen` | HGNC gene symbols | Assigned code | HUGO Gene Nomenclature Committee | 44,997 | `^[A-Za-z0-9@#\-\._]+$` |
503
+ | `bio/pdb` | PDB entry IDs | Assigned code | wwPDB / RCSB | 254,978 | `^[1-9][A-Z0-9]{3}$` |
504
 
505
  ### Classification (`cls`)
506
 
507
+ | ID | System | Kind | Authority | Codes | Pattern |
508
+ | --- | --- | --- | --- | ---: | --- |
509
+ | `cls/elm` | Chemical element symbols | Symbol | IUPAC | 118 | `^[A-Z][a-z]{0,2}$` |
510
+ | `cls/icd` | ICD-10-CM | Assigned code | US CDC / NCHS | 98,403 | `^[A-Z][A-Z0-9]{2,6}$` |
511
+ | `cls/loc` | Library of Congress classification outline | Assigned code | Library of Congress | 234 | `^[A-Z]{1,3}$` |
512
 
513
  ### Computing (`cmp`)
514
 
515
+ | ID | System | Kind | Authority | Codes | Pattern |
516
+ | --- | --- | --- | --- | ---: | --- |
517
+ | `cmp/asm` | x86 instruction mnemonics | Vocabulary term | Intel / AMD | 1,309 | `^[A-Z][A-Za-z0-9]{1,15}$` |
518
+ | `cmp/css` | CSS functions | Vocabulary term | W3C CSSWG / MDN | 105 | `^[a-zA-Z][a-zA-Z0-9\-]*$` |
519
+ | `cmp/dns` | DNS resource record types | Assigned code | IANA | 97 | `^([A-Z][A-Z0-9\-]*\|\*)$` |
520
+ | `cmp/ern` | POSIX errno names | Assigned code | IEEE / The Open Group | 77 | `^E[A-Z0-9]+$` |
521
+ | `cmp/ext` | File extensions | Community convention | Community | 286 | `^[a-z0-9][a-z0-9.+\-]*$` |
522
+ | `cmp/frc` | FourCC codec identifiers | Community convention | Community / VLC | 672 | `^[A-Za-z0-9\-.:_]{2,4}$` |
523
+ | `cmp/git` | Git subcommands | Vocabulary term | Git project | 158 | `^[a-z][a-z0-9\-]{1,17}$` |
524
+ | `cmp/hex` | CSS named colors | Vocabulary term | W3C CSSWG | 148 | `^[a-z]{3,20}$` |
525
+ | `cmp/htt` | HTTP methods | Assigned code | IANA | 41 | `^([A-Z][A-Z\-]*\|\*)$` |
526
+ | `cmp/mim` | Media types | Assigned code | IANA | 2,300 | `^[a-z]+/[a-zA-Z0-9!#$&\-\^_\.\+]+$` |
527
+ | `cmp/reg` | x86-64 register names | Vocabulary term | LLVM Project | 257 | `^[a-z][a-z0-9.()]*$` |
528
+ | `cmp/sts` | HTTP status codes | Assigned code | IANA | 500 | `^\d{3}$` |
529
+ | `cmp/vim` | Vim motion commands | Vocabulary term | Vim project | 289 | `^\S{1,40}$` |
530
 
531
  ### Communication (`com`)
532
 
533
+ | ID | System | Kind | Authority | Codes | Pattern |
534
+ | --- | --- | --- | --- | ---: | --- |
535
+ | `com/bau` | Baudot / ITA2 teleprinter codes | Symbol | ITU-T | 30 | `^[A-Z0-9\-]+$` |
536
+ | `com/mor` | Morse code | Symbol | ITU-R | 48 | `^([A-Z0-9\u00c9]\|[^\w\s]{1,2})$` |
537
+ | `com/tlg` | ABC universal commercial telegraphic code | Symbol | William Clauson-Thue (1901) | 7,359 | `^[A-Za-z]{4,22}$` |
538
 
539
  ### Finance (`fin`)
540
 
541
+ | ID | System | Kind | Authority | Codes | Pattern |
542
+ | --- | --- | --- | --- | ---: | --- |
543
+ | `fin/cfi` | ISO 10962 CFI codes | Assigned code | SIX Group / ISO | 1,216 | `^[A-Z]{6}$` |
544
+ | `fin/i42` | ISO 4217 currency codes | Assigned code | SIX Group / ISO | 178 | `^[A-Z]{3}$` |
545
+ | `fin/tck` | SEC company tickers | Assigned code | US SEC | 10,405 | `^[A-Z][A-Z\-]{0,6}$` |
546
 
547
  ### Graph and vocabulary (`grf`)
548
 
549
+ | ID | System | Kind | Authority | Codes | Pattern |
550
+ | --- | --- | --- | --- | ---: | --- |
551
+ | `grf/act` | ActivityStreams 2.0 terms | Vocabulary term | W3C | 141 | `^[A-Za-z][A-Za-z0-9]*$` |
552
+ | `grf/foa` | FOAF vocabulary terms | Vocabulary term | FOAF project | 75 | `^[A-Za-z][A-Za-z0-9_]*$` |
553
+ | `grf/gql` | openCypher keywords | Vocabulary term | openCypher / Neo4j | 63 | `^[A-Z][A-Z_]*$` |
554
+ | `grf/lpg` | Property graph schema terms | Vocabulary term | arXiv 2211.10962 | 34 | `^[A-Z]{3,9}$` |
555
+ | `grf/oid` | OpenID Connect discovery fields | Vocabulary term | OpenID Foundation | 17 | `^[a-z][a-z0-9_]*$` |
556
+ | `grf/org` | W3C Organization ontology terms | Vocabulary term | W3C | 45 | `^[a-zA-Z][A-Za-z0-9]*$` |
557
+ | `grf/rdf` | RDF Schema terms | Vocabulary term | W3C | 15 | `^[a-zA-Z][A-Za-z0-9]*$` |
558
+ | `grf/sch` | schema.org types and properties | Vocabulary term | schema.org / W3C CG | 2,987 | `^[A-Za-z0-9][A-Za-z0-9]*$` |
559
+ | `grf/shc` | SHACL vocabulary terms | Vocabulary term | W3C | 221 | `^[A-Za-z][A-Za-z0-9\-]*$` |
560
+ | `grf/sio` | SIOC vocabulary terms | Vocabulary term | SIOC Project / DERI | 99 | `^[A-Za-z][A-Za-z0-9_]*$` |
561
+ | `grf/vcd` | vCard ontology terms | Vocabulary term | W3C | 146 | `^[a-zA-Z][A-Za-z0-9\-]*$` |
562
 
563
  ### Language (`lng`)
564
 
565
+ | ID | System | Kind | Authority | Codes | Pattern |
566
+ | --- | --- | --- | --- | ---: | --- |
567
+ | `lng/b47` | BCP 47 language subtags | Assigned code | IANA | 9,295 | `^[A-Za-z0-9\-]+(\.\.[A-Za-z0-9\-]+)?$` |
568
+ | `lng/glt` | Glottolog languoid codes | Assigned code | MPI-EVA Leipzig | 27,177 | `^[a-z0-9]{4}\d{4}$` |
569
+ | `lng/i15` | ISO 15924 script codes | Assigned code | Unicode Consortium (RA) | 226 | `^[A-Z][a-z]{3}$` |
570
+ | `lng/i31` | ISO 3166-1 country codes | Assigned code | ISO 3166/MA | 249 | `^[A-Z]{2,3}$` |
571
+ | `lng/i39` | ISO 639-1 language codes | Assigned code | ISO 639/RA | 183 | `^[a-z]{2}$` |
572
+ | `lng/i63` | BCP 47 three-letter language subtags | Assigned code | IANA | 8,084 | `^[a-z]{3}$` |
573
+ | `lng/ioc` | Country codes composite | Assigned code | Community / datasets.io | 226 | `^[A-Z]{2,3}$` |
574
+ | `lng/mdy` | Month names | Vocabulary term | Unicode CLDR | 12 | `^[A-Z][a-z]{2}$` |
575
 
576
  ### Lexical (`nte`)
577
 
578
+ | ID | System | Kind | Authority | Codes | Pattern |
579
+ | --- | --- | --- | --- | ---: | --- |
580
+ | `nte/wnt` | English WordNet 2024 synset IDs | Assigned code | Global WordNet Association | 120,630 | `^oewn-\d{8}-[nvasr]$` |
581
 
582
  ### Transport and place (`trp`)
583
 
584
+ | ID | System | Kind | Authority | Codes | Pattern |
585
+ | --- | --- | --- | --- | ---: | --- |
586
+ | `trp/iata` | IATA airport codes | Assigned code | IATA | 9,056 | `^[A-Z]{3}$` |
587
+ | `trp/ica2` | IATA and ICAO airline codes | Assigned code | IATA / ICAO | 5,841 | `^[A-Z0-9]{2,3}$` |
588
+ | `trp/icao` | ICAO airport codes | Assigned code | ICAO | 10,153 | `^[A-Z]{4}$` |
589
+ | `trp/olc` | Open Location Code (plus codes) | Test vector | Google | 292 | `^[023456789CFGHJMPQRVWX]{4,}\+[23456789CFGHJMPQRVWX]*$` |
590
+ | `trp/tld` | Top level domains | Assigned code | IANA | 1,437 | `^[A-Z0-9\-]+$` |
591
+ | `trp/unl` | UN/LOCODE location codes | Assigned code | UNECE | 116,213 | `^[A-Z0-9]{3}$` |
592
 
593
  ### Writing and time (`wrt`)
594
 
595
+ | ID | System | Kind | Authority | Codes | Pattern |
596
+ | --- | --- | --- | --- | ---: | --- |
597
+ | `wrt/abj` | Hebrew script characters | Symbol | Unicode / community | 84 | `^\S$` |
598
+ | `wrt/iau` | IAU constellation abbreviations | Assigned code | International Astronomical Union | 88 | `^[A-Z][A-Za-z]{2}$` |
599
+ | `wrt/tz` | tz database zone names and abbreviations | Assigned code | IANA / tzdb | 806 | `^[A-Za-z0-9+\-_/]{1,32}$` |
600
+ | `wrt/uni` | Unicode emoji sequences | Symbol | Unicode Consortium | 5,225 | `^[0-9A-F]{4,6}( [0-9A-F]{4,6})*$` |