@mailwoman/corpus 8.6.0 → 9.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +1 -1
- package/out/src/adapter.d.ts +41 -0
- package/out/src/adapter.d.ts.map +1 -1
- package/out/src/adapter.js +46 -1
- package/out/src/adapter.js.map +1 -1
- package/out/src/adapters/ban/adapter.d.ts +4 -3
- package/out/src/adapters/ban/adapter.d.ts.map +1 -1
- package/out/src/adapters/ban/adapter.js +60 -67
- package/out/src/adapters/ban/adapter.js.map +1 -1
- package/out/src/adapters/ban/street-decompose.d.ts.map +1 -1
- package/out/src/adapters/ban/street-decompose.js +18 -25
- package/out/src/adapters/ban/street-decompose.js.map +1 -1
- package/out/src/adapters/fcc-bdc/adapter.d.ts +0 -6
- package/out/src/adapters/fcc-bdc/adapter.d.ts.map +1 -1
- package/out/src/adapters/fcc-bdc/adapter.js +2 -24
- package/out/src/adapters/fcc-bdc/adapter.js.map +1 -1
- package/out/src/adapters/geonames/adapter.d.ts.map +1 -1
- package/out/src/adapters/geonames/adapter.js +70 -76
- package/out/src/adapters/geonames/adapter.js.map +1 -1
- package/out/src/adapters/geonames-postal/adapter.d.ts.map +1 -1
- package/out/src/adapters/geonames-postal/adapter.js +44 -49
- package/out/src/adapters/geonames-postal/adapter.js.map +1 -1
- package/out/src/adapters/gnaf/adapter.d.ts.map +1 -1
- package/out/src/adapters/gnaf/adapter.js +4 -7
- package/out/src/adapters/gnaf/adapter.js.map +1 -1
- package/out/src/adapters/gnaf/assemble.d.ts.map +1 -1
- package/out/src/adapters/gnaf/assemble.js +6 -11
- package/out/src/adapters/gnaf/assemble.js.map +1 -1
- package/out/src/adapters/openaddresses/adapter.d.ts.map +1 -1
- package/out/src/adapters/openaddresses/adapter.js +2 -7
- package/out/src/adapters/openaddresses/adapter.js.map +1 -1
- package/out/src/adapters/overture/adapter.d.ts.map +1 -1
- package/out/src/adapters/overture/adapter.js +3 -7
- package/out/src/adapters/overture/adapter.js.map +1 -1
- package/out/src/adapters/state-hi-schools/adapter.d.ts.map +1 -1
- package/out/src/adapters/state-hi-schools/adapter.js +55 -73
- package/out/src/adapters/state-hi-schools/adapter.js.map +1 -1
- package/out/src/adapters/state-ia-contractors/adapter.d.ts.map +1 -1
- package/out/src/adapters/state-ia-contractors/adapter.js +58 -76
- package/out/src/adapters/state-ia-contractors/adapter.js.map +1 -1
- package/out/src/adapters/state-ny-notaries/adapter.d.ts.map +1 -1
- package/out/src/adapters/state-ny-notaries/adapter.js +67 -85
- package/out/src/adapters/state-ny-notaries/adapter.js.map +1 -1
- package/out/src/adapters/state-tx-notaries/adapter.d.ts.map +1 -1
- package/out/src/adapters/state-tx-notaries/adapter.js +69 -87
- package/out/src/adapters/state-tx-notaries/adapter.js.map +1 -1
- package/out/src/adapters/synth-po-box/adapter.d.ts.map +1 -1
- package/out/src/adapters/synth-po-box/adapter.js +7 -15
- package/out/src/adapters/synth-po-box/adapter.js.map +1 -1
- package/out/src/adapters/tiger/street-decompose.d.ts.map +1 -1
- package/out/src/adapters/tiger/street-decompose.js +18 -27
- package/out/src/adapters/tiger/street-decompose.js.map +1 -1
- package/out/src/adapters/usgov-hrsa-fqhc/adapter.d.ts +3 -3
- package/out/src/adapters/usgov-hrsa-fqhc/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-hrsa-fqhc/adapter.js +62 -80
- package/out/src/adapters/usgov-hrsa-fqhc/adapter.js.map +1 -1
- package/out/src/adapters/usgov-imls-pls/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-imls-pls/adapter.js +60 -82
- package/out/src/adapters/usgov-imls-pls/adapter.js.map +1 -1
- package/out/src/adapters/usgov-irs-bmf/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-irs-bmf/adapter.js +58 -65
- package/out/src/adapters/usgov-irs-bmf/adapter.js.map +1 -1
- package/out/src/adapters/usgov-nad/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-nad/adapter.js +5 -8
- package/out/src/adapters/usgov-nad/adapter.js.map +1 -1
- package/out/src/adapters/usgov-nppes/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-nppes/adapter.js +58 -76
- package/out/src/adapters/usgov-nppes/adapter.js.map +1 -1
- package/out/src/adapters/usgov-samhsa-treatment-locator/adapter.d.ts.map +1 -1
- package/out/src/adapters/usgov-samhsa-treatment-locator/adapter.js +51 -71
- package/out/src/adapters/usgov-samhsa-treatment-locator/adapter.js.map +1 -1
- package/out/src/adapters/wof-admin-jp/adapter.d.ts.map +1 -1
- package/out/src/adapters/wof-admin-jp/adapter.js +24 -6
- package/out/src/adapters/wof-admin-jp/adapter.js.map +1 -1
- package/out/src/build.d.ts.map +1 -1
- package/out/src/build.js +2 -1
- package/out/src/build.js.map +1 -1
- package/out/src/golden.d.ts +5 -0
- package/out/src/golden.d.ts.map +1 -1
- package/out/src/golden.js +23 -9
- package/out/src/golden.js.map +1 -1
- package/out/src/parquet-wrapper/reader.d.ts +8 -0
- package/out/src/parquet-wrapper/reader.d.ts.map +1 -1
- package/out/src/parquet-wrapper/reader.js +16 -0
- package/out/src/parquet-wrapper/reader.js.map +1 -1
- package/out/src/parquet-wrapper/schema.d.ts +4 -0
- package/out/src/parquet-wrapper/schema.d.ts.map +1 -1
- package/out/src/parquet-wrapper/schema.js.map +1 -1
- package/out/src/parquet.d.ts.map +1 -1
- package/out/src/parquet.js +2 -13
- package/out/src/parquet.js.map +1 -1
- package/out/src/shard-recipes/anchor-absorption.d.ts.map +1 -1
- package/out/src/shard-recipes/anchor-absorption.js +2 -1
- package/out/src/shard-recipes/anchor-absorption.js.map +1 -1
- package/out/src/shard-recipes/country-balanced.d.ts.map +1 -1
- package/out/src/shard-recipes/country-balanced.js +68 -60
- package/out/src/shard-recipes/country-balanced.js.map +1 -1
- package/out/src/shard-recipes/fr-fragment.d.ts.map +1 -1
- package/out/src/shard-recipes/fr-fragment.js +8 -5
- package/out/src/shard-recipes/fr-fragment.js.map +1 -1
- package/out/src/shard-recipes/fr-order.d.ts.map +1 -1
- package/out/src/shard-recipes/fr-order.js +14 -58
- package/out/src/shard-recipes/fr-order.js.map +1 -1
- package/out/src/shard-recipes/german.d.ts.map +1 -1
- package/out/src/shard-recipes/german.js +11 -58
- package/out/src/shard-recipes/german.js.map +1 -1
- package/out/src/shard-recipes/index.d.ts.map +1 -1
- package/out/src/shard-recipes/index.js +2 -0
- package/out/src/shard-recipes/index.js.map +1 -1
- package/out/src/shard-recipes/intersection.d.ts +2 -2
- package/out/src/shard-recipes/intersection.d.ts.map +1 -1
- package/out/src/shard-recipes/intersection.js +24 -62
- package/out/src/shard-recipes/intersection.js.map +1 -1
- package/out/src/shard-recipes/locale.d.ts.map +1 -1
- package/out/src/shard-recipes/locale.js +14 -16
- package/out/src/shard-recipes/locale.js.map +1 -1
- package/out/src/shard-recipes/no-fragment.d.ts.map +1 -1
- package/out/src/shard-recipes/no-fragment.js +8 -5
- package/out/src/shard-recipes/no-fragment.js.map +1 -1
- package/out/src/shard-recipes/no-street-led.d.ts.map +1 -1
- package/out/src/shard-recipes/no-street-led.js +8 -5
- package/out/src/shard-recipes/no-street-led.js.map +1 -1
- package/out/src/shard-recipes/po-box-cedex.d.ts +3 -3
- package/out/src/shard-recipes/po-box-cedex.d.ts.map +1 -1
- package/out/src/shard-recipes/po-box-cedex.js +51 -94
- package/out/src/shard-recipes/po-box-cedex.js.map +1 -1
- package/out/src/shard-recipes/scaffold.d.ts +59 -9
- package/out/src/shard-recipes/scaffold.d.ts.map +1 -1
- package/out/src/shard-recipes/scaffold.js +51 -30
- package/out/src/shard-recipes/scaffold.js.map +1 -1
- package/out/src/shard-recipes/street-affix.d.ts.map +1 -1
- package/out/src/shard-recipes/street-affix.js +56 -82
- package/out/src/shard-recipes/street-affix.js.map +1 -1
- package/out/src/shard-recipes/sub-venue-sources.d.ts +231 -0
- package/out/src/shard-recipes/sub-venue-sources.d.ts.map +1 -0
- package/out/src/shard-recipes/sub-venue-sources.js +669 -0
- package/out/src/shard-recipes/sub-venue-sources.js.map +1 -0
- package/out/src/shard-recipes/sub-venue.d.ts +269 -0
- package/out/src/shard-recipes/sub-venue.d.ts.map +1 -0
- package/out/src/shard-recipes/sub-venue.js +709 -0
- package/out/src/shard-recipes/sub-venue.js.map +1 -0
- package/out/src/shard-recipes/unit.d.ts +1 -1
- package/out/src/shard-recipes/unit.d.ts.map +1 -1
- package/out/src/shard-recipes/unit.js +22 -64
- package/out/src/shard-recipes/unit.js.map +1 -1
- package/out/src/split.d.ts +7 -0
- package/out/src/split.d.ts.map +1 -1
- package/out/src/split.js +7 -0
- package/out/src/split.js.map +1 -1
- package/out/src/synthesize.d.ts.map +1 -1
- package/out/src/synthesize.js +1 -12
- package/out/src/synthesize.js.map +1 -1
- package/out/src/tools/align-shard.d.ts.map +1 -1
- package/out/src/tools/align-shard.js +3 -7
- package/out/src/tools/align-shard.js.map +1 -1
- package/out/src/tools/audit.d.ts.map +1 -1
- package/out/src/tools/audit.js +4 -4
- package/out/src/tools/audit.js.map +1 -1
- package/out/src/tools/corpus-stats.d.ts +2 -2
- package/out/src/tools/corpus-stats.d.ts.map +1 -1
- package/out/src/tools/corpus-stats.js +18 -31
- package/out/src/tools/corpus-stats.js.map +1 -1
- package/out/src/tools/fetch/download.d.ts.map +1 -1
- package/out/src/tools/fetch/download.js +5 -6
- package/out/src/tools/fetch/download.js.map +1 -1
- package/out/src/tools/fetch/imls-pls.d.ts.map +1 -1
- package/out/src/tools/fetch/imls-pls.js +2 -2
- package/out/src/tools/fetch/imls-pls.js.map +1 -1
- package/out/src/tools/fetch/index.d.ts +14 -1
- package/out/src/tools/fetch/index.d.ts.map +1 -1
- package/out/src/tools/fetch/index.js +14 -1
- package/out/src/tools/fetch/index.js.map +1 -1
- package/out/src/tools/fetch/nad.d.ts +1 -1
- package/out/src/tools/fetch/nad.js +1 -1
- package/out/src/tools/fetch/nppes.d.ts.map +1 -1
- package/out/src/tools/fetch/nppes.js +2 -1
- package/out/src/tools/fetch/nppes.js.map +1 -1
- package/out/src/tools/fetch/openaddresses.d.ts +1 -1
- package/out/src/tools/fetch/openaddresses.js +1 -1
- package/out/src/tools/fetch/ourairports.d.ts +45 -0
- package/out/src/tools/fetch/ourairports.d.ts.map +1 -0
- package/out/src/tools/fetch/ourairports.js +123 -0
- package/out/src/tools/fetch/ourairports.js.map +1 -0
- package/out/src/tools/fetch/wikidata-subvenue.d.ts +171 -0
- package/out/src/tools/fetch/wikidata-subvenue.d.ts.map +1 -0
- package/out/src/tools/fetch/wikidata-subvenue.js +275 -0
- package/out/src/tools/fetch/wikidata-subvenue.js.map +1 -0
- package/out/src/tools/golden-expand.d.ts.map +1 -1
- package/out/src/tools/golden-expand.js +18 -16
- package/out/src/tools/golden-expand.js.map +1 -1
- package/out/src/tools/golden-promote.d.ts.map +1 -1
- package/out/src/tools/golden-promote.js +12 -6
- package/out/src/tools/golden-promote.js.map +1 -1
- package/out/src/tools/golden-relabel-street.d.ts +196 -0
- package/out/src/tools/golden-relabel-street.d.ts.map +1 -0
- package/out/src/tools/golden-relabel-street.js +513 -0
- package/out/src/tools/golden-relabel-street.js.map +1 -0
- package/out/src/tools/index.d.ts +4 -0
- package/out/src/tools/index.d.ts.map +1 -1
- package/out/src/tools/index.js +4 -0
- package/out/src/tools/index.js.map +1 -1
- package/out/src/tools/jsonl-to-parquet.d.ts.map +1 -1
- package/out/src/tools/jsonl-to-parquet.js +3 -2
- package/out/src/tools/jsonl-to-parquet.js.map +1 -1
- package/out/src/tools/lint-shard-vocab.d.ts.map +1 -1
- package/out/src/tools/lint-shard-vocab.js +3 -2
- package/out/src/tools/lint-shard-vocab.js.map +1 -1
- package/out/src/tools/lint-shard.d.ts +3 -3
- package/out/src/tools/lint-shard.d.ts.map +1 -1
- package/out/src/tools/lint-shard.js +23 -32
- package/out/src/tools/lint-shard.js.map +1 -1
- package/out/src/tools/overlay-manifest.d.ts.map +1 -1
- package/out/src/tools/overlay-manifest.js +2 -1
- package/out/src/tools/overlay-manifest.js.map +1 -1
- package/out/src/tools/overture-subvenue.d.ts +112 -0
- package/out/src/tools/overture-subvenue.d.ts.map +1 -0
- package/out/src/tools/overture-subvenue.js +144 -0
- package/out/src/tools/overture-subvenue.js.map +1 -0
- package/out/src/tools/shard-kryptonite.d.ts +1 -1
- package/out/src/tools/shard-kryptonite.d.ts.map +1 -1
- package/out/src/tools/shard-kryptonite.js +5 -4
- package/out/src/tools/shard-kryptonite.js.map +1 -1
- package/out/src/tools/shard-translit.d.ts +6 -3
- package/out/src/tools/shard-translit.d.ts.map +1 -1
- package/out/src/tools/shard-translit.js +8 -6
- package/out/src/tools/shard-translit.js.map +1 -1
- package/out/src/tools/sub-venue-lexicon.d.ts +507 -0
- package/out/src/tools/sub-venue-lexicon.d.ts.map +1 -0
- package/out/src/tools/sub-venue-lexicon.js +817 -0
- package/out/src/tools/sub-venue-lexicon.js.map +1 -0
- package/out/src/tools/sub-venue-promotions.d.ts +94 -0
- package/out/src/tools/sub-venue-promotions.d.ts.map +1 -0
- package/out/src/tools/sub-venue-promotions.js +266 -0
- package/out/src/tools/sub-venue-promotions.js.map +1 -0
- package/out/src/wof-json.d.ts +2 -2
- package/out/src/wof-json.d.ts.map +1 -1
- package/out/src/wof-json.js +5 -8
- package/out/src/wof-json.js.map +1 -1
- package/package.json +8 -8
- package/src/adapter.ts +58 -1
- package/src/adapters/ban/adapter.ts +55 -65
- package/src/adapters/ban/street-decompose.ts +17 -25
- package/src/adapters/fcc-bdc/adapter.ts +2 -33
- package/src/adapters/geonames/adapter.ts +64 -72
- package/src/adapters/geonames-postal/adapter.ts +42 -50
- package/src/adapters/gnaf/adapter.ts +4 -7
- package/src/adapters/gnaf/assemble.ts +6 -10
- package/src/adapters/openaddresses/adapter.ts +2 -7
- package/src/adapters/overture/adapter.ts +3 -6
- package/src/adapters/state-hi-schools/adapter.ts +50 -73
- package/src/adapters/state-ia-contractors/adapter.ts +52 -76
- package/src/adapters/state-ny-notaries/adapter.ts +61 -85
- package/src/adapters/state-tx-notaries/adapter.ts +60 -84
- package/src/adapters/synth-po-box/adapter.ts +7 -17
- package/src/adapters/tiger/street-decompose.ts +21 -31
- package/src/adapters/usgov-hrsa-fqhc/adapter.ts +61 -84
- package/src/adapters/usgov-imls-pls/adapter.ts +54 -82
- package/src/adapters/usgov-irs-bmf/adapter.ts +53 -63
- package/src/adapters/usgov-nad/adapter.ts +5 -8
- package/src/adapters/usgov-nppes/adapter.ts +51 -75
- package/src/adapters/usgov-samhsa-treatment-locator/adapter.ts +45 -71
- package/src/adapters/wof-admin-jp/adapter.ts +34 -8
- package/src/build.ts +2 -1
- package/src/golden.ts +24 -9
- package/src/parquet-wrapper/reader.ts +19 -0
- package/src/parquet-wrapper/schema.ts +4 -0
- package/src/parquet.ts +3 -16
- package/src/shard-recipes/anchor-absorption.ts +2 -1
- package/src/shard-recipes/country-balanced.ts +69 -70
- package/src/shard-recipes/fr-fragment.ts +10 -7
- package/src/shard-recipes/fr-order.ts +13 -63
- package/src/shard-recipes/german.ts +12 -65
- package/src/shard-recipes/index.ts +2 -0
- package/src/shard-recipes/intersection.ts +32 -60
- package/src/shard-recipes/locale.ts +14 -16
- package/src/shard-recipes/no-fragment.ts +10 -7
- package/src/shard-recipes/no-street-led.ts +10 -7
- package/src/shard-recipes/po-box-cedex.ts +68 -119
- package/src/shard-recipes/scaffold.ts +82 -28
- package/src/shard-recipes/street-affix.ts +58 -95
- package/src/shard-recipes/sub-venue-sources.ts +895 -0
- package/src/shard-recipes/sub-venue.ts +982 -0
- package/src/shard-recipes/unit.ts +22 -70
- package/src/split.ts +7 -0
- package/src/synthesize.ts +1 -15
- package/src/tools/align-shard.ts +3 -6
- package/src/tools/audit.ts +6 -5
- package/src/tools/corpus-stats.ts +25 -33
- package/src/tools/fetch/download.ts +7 -5
- package/src/tools/fetch/imls-pls.ts +2 -2
- package/src/tools/fetch/index.ts +14 -1
- package/src/tools/fetch/nad.ts +1 -1
- package/src/tools/fetch/nppes.ts +2 -1
- package/src/tools/fetch/openaddresses.ts +1 -1
- package/src/tools/fetch/ourairports.ts +166 -0
- package/src/tools/fetch/wikidata-subvenue.ts +386 -0
- package/src/tools/golden-expand.ts +18 -13
- package/src/tools/golden-promote.ts +15 -6
- package/src/tools/golden-relabel-street.ts +748 -0
- package/src/tools/index.ts +4 -0
- package/src/tools/jsonl-to-parquet.ts +3 -2
- package/src/tools/lint-shard-vocab.ts +3 -2
- package/src/tools/lint-shard.ts +33 -36
- package/src/tools/overlay-manifest.ts +2 -1
- package/src/tools/overture-subvenue.ts +215 -0
- package/src/tools/shard-kryptonite.ts +5 -4
- package/src/tools/shard-translit.ts +12 -7
- package/src/tools/sub-venue-lexicon.ts +1250 -0
- package/src/tools/sub-venue-promotions.ts +330 -0
- package/src/wof-json.ts +5 -8
|
@@ -14,9 +14,8 @@
|
|
|
14
14
|
*
|
|
15
15
|
* License: stamped `"Public Domain"` per Texas state government open-data terms.
|
|
16
16
|
*/
|
|
17
|
-
import {
|
|
18
|
-
import {
|
|
19
|
-
import { stableSourceID } from "../../adapter.js";
|
|
17
|
+
import { CSVSpliterator } from "spliterator";
|
|
18
|
+
import { splitStreetLine, stableSourceID } from "../../adapter.js";
|
|
20
19
|
import { lookupStateAbbreviation } from "../../codex/us-fips-state.js";
|
|
21
20
|
import { reconcileComponents } from "../../format.js";
|
|
22
21
|
/**
|
|
@@ -29,20 +28,10 @@ export const STATE_TX_NOTARIES_ADAPTER_ID = "state-tx-notaries";
|
|
|
29
28
|
* than having to look them up.
|
|
30
29
|
*/
|
|
31
30
|
export const STATE_TX_NOTARIES_DEFAULT_LICENSE = "Public Domain";
|
|
32
|
-
const HOUSE_NUMBER_PREFIX = /^(\d+(?:-\d+)?[A-Za-z]?)\s+(.+)$/;
|
|
33
31
|
/**
|
|
34
32
|
* Match trailing "CITY, ST ZIP" or "CITY, ST" at the end of an address line.
|
|
35
33
|
*/
|
|
36
34
|
const CITY_STATE_ZIP_SUFFIX = /[,]?\s*([^,]+),\s*([A-Z]{2})\s*(\d{5}(?:-\d{4})?)?\s*$/i;
|
|
37
|
-
function splitAddress(address) {
|
|
38
|
-
const trimmed = address.trim();
|
|
39
|
-
if (!trimmed)
|
|
40
|
-
return null;
|
|
41
|
-
const m = HOUSE_NUMBER_PREFIX.exec(trimmed);
|
|
42
|
-
if (m)
|
|
43
|
-
return { house_number: m[1], street: m[2].trim() };
|
|
44
|
-
return { street: trimmed };
|
|
45
|
-
}
|
|
46
35
|
export function createStateTxNotariesAdapter() {
|
|
47
36
|
return {
|
|
48
37
|
id: STATE_TX_NOTARIES_ADAPTER_ID,
|
|
@@ -52,81 +41,74 @@ export function createStateTxNotariesAdapter() {
|
|
|
52
41
|
if (opts.country && opts.country !== "US") {
|
|
53
42
|
throw new Error(`state-tx-notaries adapter: only US supported, got country=${opts.country}`);
|
|
54
43
|
}
|
|
55
|
-
const
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
relax_column_count: true,
|
|
61
|
-
}));
|
|
44
|
+
const rows = CSVSpliterator.fromAsync(opts.inputPath, {
|
|
45
|
+
mode: "object",
|
|
46
|
+
normalizeKeys: false,
|
|
47
|
+
enableQuoteHandling: true,
|
|
48
|
+
});
|
|
62
49
|
let emitted = 0;
|
|
63
|
-
|
|
64
|
-
|
|
65
|
-
|
|
66
|
-
|
|
67
|
-
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
77
|
-
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
82
|
-
|
|
83
|
-
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
87
|
-
|
|
88
|
-
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
|
|
92
|
-
|
|
93
|
-
|
|
94
|
-
|
|
95
|
-
|
|
96
|
-
|
|
97
|
-
|
|
98
|
-
|
|
99
|
-
|
|
100
|
-
|
|
101
|
-
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
|
|
111
|
-
|
|
112
|
-
|
|
113
|
-
|
|
114
|
-
|
|
115
|
-
|
|
116
|
-
|
|
117
|
-
|
|
118
|
-
|
|
119
|
-
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
|
|
123
|
-
|
|
124
|
-
|
|
125
|
-
emitted++;
|
|
126
|
-
}
|
|
127
|
-
}
|
|
128
|
-
finally {
|
|
129
|
-
stream.destroy();
|
|
50
|
+
for await (const record of rows) {
|
|
51
|
+
if (opts.signal?.aborted)
|
|
52
|
+
break;
|
|
53
|
+
if (opts.limit !== undefined && emitted >= opts.limit)
|
|
54
|
+
break;
|
|
55
|
+
const rawAddress = (record.Address ?? "").trim();
|
|
56
|
+
if (!rawAddress)
|
|
57
|
+
continue;
|
|
58
|
+
const firstName = (record["First Name"] ?? "").trim();
|
|
59
|
+
const lastName = (record["Last Name"] ?? "").trim();
|
|
60
|
+
const notaryID = (record["Notary ID"] ?? "").trim();
|
|
61
|
+
// Parse embedded city/state/zip from the trailing portion of the address.
|
|
62
|
+
// Addresses look like: "1215 MCMILLAN DR\nCEDAR HILL, TX 75104"
|
|
63
|
+
const addrSingleLine = rawAddress.replaceAll("\n", ", ");
|
|
64
|
+
const cszMatch = CITY_STATE_ZIP_SUFFIX.exec(addrSingleLine);
|
|
65
|
+
if (!cszMatch)
|
|
66
|
+
continue;
|
|
67
|
+
const city = (cszMatch[1] ?? "").trim();
|
|
68
|
+
const stateAbbr = (cszMatch[2] ?? "").trim();
|
|
69
|
+
const zip = (cszMatch[3] ?? "").trim();
|
|
70
|
+
if (!city || !stateAbbr)
|
|
71
|
+
continue;
|
|
72
|
+
const state = lookupStateAbbreviation(stateAbbr);
|
|
73
|
+
if (!state)
|
|
74
|
+
continue;
|
|
75
|
+
// Extract the street portion (everything before the city/state/zip)
|
|
76
|
+
const streetPortion = addrSingleLine.slice(0, cszMatch.index).replace(/,\s*$/, "").trim();
|
|
77
|
+
if (!streetPortion)
|
|
78
|
+
continue;
|
|
79
|
+
const split = splitStreetLine(streetPortion);
|
|
80
|
+
if (!split)
|
|
81
|
+
continue;
|
|
82
|
+
const venue = [firstName, lastName].filter(Boolean).join(" ") || undefined;
|
|
83
|
+
const components = {
|
|
84
|
+
...(venue ? { venue } : {}),
|
|
85
|
+
...(split.house_number ? { house_number: split.house_number } : {}),
|
|
86
|
+
street: split.street,
|
|
87
|
+
locality: city,
|
|
88
|
+
region: state.abbreviation,
|
|
89
|
+
...(zip ? { postcode: zip } : {}),
|
|
90
|
+
};
|
|
91
|
+
const streetPart = [split.house_number, split.street].filter(Boolean).join(" ").trim();
|
|
92
|
+
const raw = [venue, streetPart, [city, [stateAbbr, zip].filter(Boolean).join(" ")].filter(Boolean).join(", ")]
|
|
93
|
+
.filter(Boolean)
|
|
94
|
+
.join(", ");
|
|
95
|
+
const aligned = reconcileComponents(components, raw);
|
|
96
|
+
if (Object.keys(aligned).length <= 2)
|
|
97
|
+
continue;
|
|
98
|
+
const sourceID = notaryID
|
|
99
|
+
? `${STATE_TX_NOTARIES_ADAPTER_ID}-${notaryID}`
|
|
100
|
+
: stableSourceID(STATE_TX_NOTARIES_ADAPTER_ID, aligned);
|
|
101
|
+
yield {
|
|
102
|
+
raw,
|
|
103
|
+
components: aligned,
|
|
104
|
+
country: "US",
|
|
105
|
+
locale: "en-US",
|
|
106
|
+
source: STATE_TX_NOTARIES_ADAPTER_ID,
|
|
107
|
+
source_id: sourceID,
|
|
108
|
+
corpus_version: "",
|
|
109
|
+
license: STATE_TX_NOTARIES_DEFAULT_LICENSE,
|
|
110
|
+
};
|
|
111
|
+
emitted++;
|
|
130
112
|
}
|
|
131
113
|
},
|
|
132
114
|
};
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/state-tx-notaries/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAEH,OAAO,EAAE,
|
|
1
|
+
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/state-tx-notaries/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAA;AAE5C,OAAO,EAAE,eAAe,EAAE,cAAc,EAAE,MAAM,kBAAkB,CAAA;AAClE,OAAO,EAAE,uBAAuB,EAAE,MAAM,8BAA8B,CAAA;AACtE,OAAO,EAAE,mBAAmB,EAAE,MAAM,iBAAiB,CAAA;AAGrD;;;GAGG;AACH,MAAM,CAAC,MAAM,4BAA4B,GAAG,mBAAmB,CAAA;AAC/D;;;GAGG;AACH,MAAM,CAAC,MAAM,iCAAiC,GAAG,eAAe,CAAA;AAEhE;;GAEG;AACH,MAAM,qBAAqB,GAAG,yDAAyD,CAAA;AASvF,MAAM,UAAU,4BAA4B;IAC3C,OAAO;QACN,EAAE,EAAE,4BAA4B;QAChC,cAAc,EAAE,iCAAiC;QACjD,WAAW,EACV,wGAAwG;QAEzG,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,KAAK,CAAC,6DAA6D,IAAI,CAAC,OAAO,EAAE,CAAC,CAAA;YAC7F,CAAC;YAED,MAAM,IAAI,GAAG,cAAc,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,EAAE;gBACrD,IAAI,EAAE,QAAQ;gBACd,aAAa,EAAE,KAAK;gBACpB,mBAAmB,EAAE,IAAI;aACzB,CAAC,CAAA;YAEF,IAAI,OAAO,GAAG,CAAC,CAAA;YAEf,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,IAAkC,EAAE,CAAC;gBAC/D,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAE/B,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;oBAAE,MAAK;gBAE5D,MAAM,UAAU,GAAG,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAEhD,IAAI,CAAC,UAAU;oBAAE,SAAQ;gBAEzB,MAAM,SAAS,GAAG,CAAC,MAAM,CAAC,YAAY,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBACrD,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBACnD,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAEnD,0EAA0E;gBAC1E,gEAAgE;gBAChE,MAAM,cAAc,GAAG,UAAU,CAAC,UAAU,CAAC,IAAI,EAAE,IAAI,CAAC,CAAA;gBACxD,MAAM,QAAQ,GAAG,qBAAqB,CAAC,IAAI,CAAC,cAAc,CAAC,CAAA;gBAE3D,IAAI,CAAC,QAAQ;oBAAE,SAAQ;gBAEvB,MAAM,IAAI,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBACvC,MAAM,SAAS,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAC5C,MAAM,GAAG,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAEtC,IAAI,CAAC,IAAI,IAAI,CAAC,SAAS;oBAAE,SAAQ;gBAEjC,MAAM,KAAK,GAAG,uBAAuB,CAAC,SAAS,CAAC,CAAA;gBAEhD,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,oEAAoE;gBACpE,MAAM,aAAa,GAAG,cAAc,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAEzF,IAAI,CAAC,aAAa;oBAAE,SAAQ;gBAE5B,MAAM,KAAK,GAAG,eAAe,CAAC,aAAa,CAAC,CAAA;gBAE5C,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,MAAM,KAAK,GAAG,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,SAAS,CAAA;gBAE1E,MAAM,UAAU,GAA+B;oBAC9C,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBAC3B,GAAG,CAAC,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,KAAK,CAAC,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACnE,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,QAAQ,EAAE,IAAI;oBACd,MAAM,EAAE,KAAK,CAAC,YAAY;oBAC1B,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;iBACjC,CAAA;gBAED,MAAM,UAAU,GAAG,CAAC,KAAK,CAAC,YAAY,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;gBAEtF,MAAM,GAAG,GAAG,CAAC,KAAK,EAAE,UAAU,EAAE,CAAC,IAAI,EAAE,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;qBAC5G,MAAM,CAAC,OAAO,CAAC;qBACf,IAAI,CAAC,IAAI,CAAC,CAAA;gBAEZ,MAAM,OAAO,GAAG,mBAAmB,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;gBAEpD,IAAI,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM,IAAI,CAAC;oBAAE,SAAQ;gBAE9C,MAAM,QAAQ,GAAG,QAAQ;oBACxB,CAAC,CAAC,GAAG,4BAA4B,IAAI,QAAQ,EAAE;oBAC/C,CAAC,CAAC,cAAc,CAAC,4BAA4B,EAAE,OAAO,CAAC,CAAA;gBAExD,MAAM;oBACL,GAAG;oBACH,UAAU,EAAE,OAAO;oBACnB,OAAO,EAAE,IAAI;oBACb,MAAM,EAAE,OAAO;oBACf,MAAM,EAAE,4BAA4B;oBACpC,SAAS,EAAE,QAAQ;oBACnB,cAAc,EAAE,EAAE;oBAClB,OAAO,EAAE,iCAAiC;iBAC1C,CAAA;gBAED,OAAO,EAAE,CAAA;YACV,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,sBAAsB,GAAG,4BAA4B,EAAE,CAAA"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;
|
|
1
|
+
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAOH,OAAO,EAKN,KAAK,cAAc,EACnB,MAAM,4BAA4B,CAAA;AACnC,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAEjF;;;GAGG;AACH,eAAO,MAAM,uBAAuB,iBAAiB,CAAA;AACrD;;;GAGG;AACH,eAAO,MAAM,oBAAoB,qEAAgE,CAAA;AAEjG,MAAM,WAAW,aAAc,SAAQ,cAAc;IACpD,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,WAAW,CAAC,EAAE,MAAM,CAAA;CACpB;AAED,MAAM,WAAW,wBAAwB;IACxC;;;OAGG;IACH,gBAAgB,CAAC,EAAE,MAAM,CAAA;IACzB;;;OAGG;IACH,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB;;OAEG;IACH,IAAI,CAAC,EAAE,MAAM,CAAA;IACb;;;;;;OAMG;IACH,aAAa,CAAC,EAAE,MAAM,CAAA;CACtB;AAED,wBAAgB,uBAAuB,CAAC,IAAI,GAAE,wBAA6B,GAAG,aAAa,CAyH1F;AAED;;GAEG;AACH,eAAO,MAAM,iBAAiB,eAA4B,CAAA"}
|
|
@@ -19,6 +19,8 @@
|
|
|
19
19
|
* ARE valid (CMRA addresses). Those are produced when `pmbRatio > 0` AND the input tuple
|
|
20
20
|
* carries a `street` field.
|
|
21
21
|
*/
|
|
22
|
+
import { tryParsingJSON } from "@mailwoman/core/objects";
|
|
23
|
+
import { makeLcg } from "@mailwoman/core/utils";
|
|
22
24
|
import { TextSpliterator } from "spliterator";
|
|
23
25
|
import { stableSourceID } from "../../adapter.js";
|
|
24
26
|
import { countryToLocale, REGION_OPTIONAL_LOCALES, synthesizeMilitaryPoBoxRow, synthesizePoBoxRow, } from "../../synthesize-po-box.js";
|
|
@@ -32,13 +34,6 @@ export const SYNTH_PO_BOX_ADAPTER_ID = "synth-po-box";
|
|
|
32
34
|
* derived from, so the attribution travels with it.
|
|
33
35
|
*/
|
|
34
36
|
export const SYNTH_PO_BOX_LICENSE = "Synthetic — derived from CC-BY / public-domain input tuples";
|
|
35
|
-
function makeRandom(seed) {
|
|
36
|
-
let s = seed;
|
|
37
|
-
return () => {
|
|
38
|
-
s = (s * 1_664_525 + 1_013_904_223) % 4_294_967_296;
|
|
39
|
-
return s / 4_294_967_296;
|
|
40
|
-
};
|
|
41
|
-
}
|
|
42
37
|
export function createSynthPoBoxAdapter(opts = {}) {
|
|
43
38
|
const variantsPerInput = opts.variantsPerInput ?? 1;
|
|
44
39
|
const pmbRatio = opts.pmbRatio ?? 0.15;
|
|
@@ -48,9 +43,9 @@ export function createSynthPoBoxAdapter(opts = {}) {
|
|
|
48
43
|
defaultLicense: SYNTH_PO_BOX_LICENSE,
|
|
49
44
|
description: "Synthetic PO box / PMB / Apartado / Boîte Postale rows. Consumes JSONL of (locality, region, postcode, country) tuples and emits locale-appropriate PO box variants.",
|
|
50
45
|
async *rows(options) {
|
|
51
|
-
const random =
|
|
52
|
-
// TextSpliterator streams string lines; the per-line
|
|
53
|
-
// tolerant of malformed rows (skipped++), so TextSpliterator +
|
|
46
|
+
const random = makeLcg(opts.seed ?? Date.now());
|
|
47
|
+
// TextSpliterator streams string lines; the per-line tryParsingJSON below keeps this reader
|
|
48
|
+
// tolerant of malformed rows (skipped++), so TextSpliterator + a non-throwing parse — not
|
|
54
49
|
// JSONSpliterator, which would throw on the first bad line.
|
|
55
50
|
const lines = TextSpliterator.fromAsync(options.inputPath);
|
|
56
51
|
let emitted = 0;
|
|
@@ -64,11 +59,8 @@ export function createSynthPoBoxAdapter(opts = {}) {
|
|
|
64
59
|
const trimmed = line.trim();
|
|
65
60
|
if (!trimmed)
|
|
66
61
|
continue;
|
|
67
|
-
|
|
68
|
-
|
|
69
|
-
input = JSON.parse(trimmed);
|
|
70
|
-
}
|
|
71
|
-
catch {
|
|
62
|
+
const input = tryParsingJSON(trimmed);
|
|
63
|
+
if (input === null) {
|
|
72
64
|
skipped++;
|
|
73
65
|
continue;
|
|
74
66
|
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAEH,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAE7C,OAAO,EAAE,cAAc,EAAE,MAAM,kBAAkB,CAAA;AACjD,OAAO,EACN,eAAe,EACf,uBAAuB,EACvB,0BAA0B,EAC1B,kBAAkB,GAElB,MAAM,4BAA4B,CAAA;AAGnC;;;GAGG;AACH,MAAM,CAAC,MAAM,uBAAuB,GAAG,cAAc,CAAA;AACrD;;;GAGG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,6DAA6D,CAAA;AAgCjG,
|
|
1
|
+
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,yBAAyB,CAAA;AACxD,OAAO,EAAE,OAAO,EAAE,MAAM,uBAAuB,CAAA;AAC/C,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAE7C,OAAO,EAAE,cAAc,EAAE,MAAM,kBAAkB,CAAA;AACjD,OAAO,EACN,eAAe,EACf,uBAAuB,EACvB,0BAA0B,EAC1B,kBAAkB,GAElB,MAAM,4BAA4B,CAAA;AAGnC;;;GAGG;AACH,MAAM,CAAC,MAAM,uBAAuB,GAAG,cAAc,CAAA;AACrD;;;GAGG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,6DAA6D,CAAA;AAgCjG,MAAM,UAAU,uBAAuB,CAAC,OAAiC,EAAE;IAC1E,MAAM,gBAAgB,GAAG,IAAI,CAAC,gBAAgB,IAAI,CAAC,CAAA;IACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAA;IACtC,MAAM,aAAa,GAAG,IAAI,CAAC,aAAa,IAAI,CAAC,CAAA;IAE7C,OAAO;QACN,EAAE,EAAE,uBAAuB;QAC3B,cAAc,EAAE,oBAAoB;QACpC,WAAW,EACV,sKAAsK;QAEvK,KAAK,CAAC,CAAC,IAAI,CAAC,OAAuB;YAClC,MAAM,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;YAE/C,4FAA4F;YAC5F,0FAA0F;YAC1F,4DAA4D;YAC5D,MAAM,KAAK,GAAG,eAAe,CAAC,SAAS,CAAC,OAAO,CAAC,SAAS,CAAC,CAAA;YAE1D,IAAI,OAAO,GAAG,CAAC,CAAA;YACf,IAAI,OAAO,GAAG,CAAC,CAAA;YACf,IAAI,WAAW,GAAG,CAAC,CAAA;YAEnB,IAAI,KAAK,EAAE,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;gBAChC,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAElC,IAAI,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,OAAO,CAAC,KAAK;oBAAE,MAAK;gBAElE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAA;gBAE3B,IAAI,CAAC,OAAO;oBAAE,SAAQ;gBAEtB,MAAM,KAAK,GAAG,cAAc,CAAgB,OAAO,CAAC,CAAA;gBAEpD,IAAI,KAAK,KAAK,IAAI,EAAE,CAAC;oBACpB,OAAO,EAAE,CAAA;oBAET,SAAQ;gBACT,CAAC;gBAED,yFAAyF;gBACzF,wFAAwF;gBACxF,qDAAqD;gBACrD,MAAM,cAAc,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,uBAAuB,CAAC,GAAG,CAAC,eAAe,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAA;gBAE1G,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,cAAc,CAAC,EAAE,CAAC;oBAChG,OAAO,EAAE,CAAA;oBAET,SAAQ;gBACT,CAAC;gBAED,IAAI,OAAO,CAAC,OAAO,IAAI,OAAO,CAAC,OAAO,KAAK,KAAK,CAAC,OAAO;oBAAE,SAAQ;gBAElE,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,gBAAgB,EAAE,CAAC,EAAE,EAAE,CAAC;oBAC3C,MAAM,KAAK,GAAG,kBAAkB,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,QAAQ,EAAE,CAAC,CAAA;oBAE7D,IAAI,CAAC,KAAK;wBAAE,SAAQ;oBAEpB,6EAA6E;oBAC7E,iDAAiD;oBACjD,MAAM,QAAQ,GAAG,cAAc,CAAC,uBAAuB,EAAE;wBACxD,QAAQ,EAAE,GAAG,KAAK,CAAC,QAAQ,IAAI,CAAC,EAAE;wBAClC,MAAM,EAAE,KAAK,CAAC,MAAM;wBACpB,QAAQ,EAAE,KAAK,CAAC,QAAQ;wBACxB,OAAO,EAAE,KAAK,CAAC,OAAO;qBACtB,CAAC,CAAA;oBAEF,MAAM;wBACL,GAAG,EAAE,KAAK,CAAC,GAAG;wBACd,UAAU,EAAE,KAAK,CAAC,UAAU;wBAC5B,OAAO,EAAE,KAAK,CAAC,OAAO;wBACtB,MAAM,EAAE,KAAK,CAAC,MAAM;wBACpB,MAAM,EAAE,uBAAuB;wBAC/B,SAAS,EAAE,QAAQ;wBACnB,cAAc,EAAE,EAAE;wBAClB,OAAO,EAAE,oBAAoB;qBAC7B,CAAA;oBAED,OAAO,EAAE,CAAA;oBAET,IAAI,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,OAAO,CAAC,KAAK;wBAAE,MAAK;gBACnE,CAAC;gBAED,0FAA0F;gBAC1F,0FAA0F;gBAC1F,wFAAwF;gBACxF,2FAA2F;gBAC3F,kDAAkD;gBAClD,MAAM,eAAe,GAAG,CAAC,OAAO,CAAC,OAAO,IAAI,OAAO,CAAC,OAAO,KAAK,IAAI,CAAA;gBAEpE,IACC,aAAa,GAAG,CAAC;oBACjB,eAAe;oBACf,CAAC,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC;oBACxD,MAAM,EAAE,GAAG,aAAa,EACvB,CAAC;oBACF,MAAM,GAAG,GAAG,0BAA0B,CAAC,EAAE,MAAM,EAAE,CAAC,CAAA;oBAElD,MAAM,QAAQ,GAAG,cAAc,CAAC,uBAAuB,EAAE;wBACxD,MAAM,EAAE,GAAG,GAAG,CAAC,UAAU,CAAC,MAAM,OAAO,WAAW,EAAE,EAAE;wBACtD,QAAQ,EAAE,GAAG,CAAC,UAAU,CAAC,QAAS;wBAClC,MAAM,EAAE,GAAG,CAAC,UAAU,CAAC,MAAO;wBAC9B,QAAQ,EAAE,GAAG,CAAC,UAAU,CAAC,QAAS;qBAClC,CAAC,CAAA;oBAEF,MAAM;wBACL,GAAG,EAAE,GAAG,CAAC,GAAG;wBACZ,UAAU,EAAE,GAAG,CAAC,UAAU;wBAC1B,OAAO,EAAE,IAAI;wBACb,MAAM,EAAE,GAAG,CAAC,MAAM;wBAClB,MAAM,EAAE,uBAAuB;wBAC/B,SAAS,EAAE,QAAQ;wBACnB,cAAc,EAAE,EAAE;wBAClB,OAAO,EAAE,oBAAoB;qBAC7B,CAAA;oBAED,OAAO,EAAE,CAAA;gBACV,CAAC;YACF,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,iBAAiB,GAAG,uBAAuB,EAAE,CAAA"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"street-decompose.d.ts","sourceRoot":"","sources":["../../../../src/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;
|
|
1
|
+
{"version":3,"file":"street-decompose.d.ts","sourceRoot":"","sources":["../../../../src/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAqCH,MAAM,WAAW,gBAAgB;IAChC,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,MAAM,EAAE,MAAM,CAAA;IACd,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;CACrB;AAED;;;;;GAKG;AACH,wBAAgB,eAAe,CAAC,QAAQ,EAAE,MAAM,GAAG,gBAAgB,CA6ClE"}
|
|
@@ -16,38 +16,29 @@
|
|
|
16
16
|
* suffix: "Blvd" }
|
|
17
17
|
*/
|
|
18
18
|
import { readFileSync } from "node:fs";
|
|
19
|
-
import { resolve } from "node:path";
|
|
20
19
|
import { resourceDictionaryPath } from "@mailwoman/core/utils";
|
|
20
|
+
import { TextSpliterator } from "spliterator";
|
|
21
21
|
function loadDictionary(filename) {
|
|
22
|
-
//
|
|
23
|
-
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
|
|
30
|
-
|
|
31
|
-
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
|
|
35
|
-
|
|
36
|
-
|
|
37
|
-
|
|
38
|
-
const f = form.trim().toLowerCase();
|
|
39
|
-
if (f) {
|
|
40
|
-
set.add(f);
|
|
41
|
-
}
|
|
42
|
-
}
|
|
22
|
+
// `resourceDictionaryPath` already resolves both layouts — `core/data/...` from source and from the
|
|
23
|
+
// packaged `out/` tree. The candidate list this replaced named it TWICE and then guessed a third path
|
|
24
|
+
// off `process.cwd()`, and swallowed every error while probing, so a corrupt dictionary reported as a
|
|
25
|
+
// missing one.
|
|
26
|
+
const text = readFileSync(resourceDictionaryPath("libpostal", "en", filename), "utf8");
|
|
27
|
+
const set = new Set();
|
|
28
|
+
// The largest libpostal dictionary is 8.4 KB, and this runs once per process at module load.
|
|
29
|
+
for (const line of TextSpliterator.from(text)) {
|
|
30
|
+
const trimmed = line.trim();
|
|
31
|
+
if (!trimmed || trimmed.startsWith("#"))
|
|
32
|
+
continue;
|
|
33
|
+
// libpostal format: canonical|abbr|abbr|... — index all forms
|
|
34
|
+
for (const form of trimmed.split("|")) {
|
|
35
|
+
const f = form.trim().toLowerCase();
|
|
36
|
+
if (f) {
|
|
37
|
+
set.add(f);
|
|
43
38
|
}
|
|
44
|
-
return set;
|
|
45
|
-
}
|
|
46
|
-
catch {
|
|
47
|
-
// try next candidate
|
|
48
39
|
}
|
|
49
40
|
}
|
|
50
|
-
|
|
41
|
+
return set;
|
|
51
42
|
}
|
|
52
43
|
const DIRECTIONALS = loadDictionary("directionals.txt");
|
|
53
44
|
const STREET_TYPES = loadDictionary("street_types.txt");
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"street-decompose.js","sourceRoot":"","sources":["../../../../src/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,YAAY,EAAE,MAAM,SAAS,CAAA;
|
|
1
|
+
{"version":3,"file":"street-decompose.js","sourceRoot":"","sources":["../../../../src/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,YAAY,EAAE,MAAM,SAAS,CAAA;AAEtC,OAAO,EAAE,sBAAsB,EAAE,MAAM,uBAAuB,CAAA;AAC9D,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAE7C,SAAS,cAAc,CAAC,QAAgB;IACvC,oGAAoG;IACpG,sGAAsG;IACtG,sGAAsG;IACtG,eAAe;IACf,MAAM,IAAI,GAAG,YAAY,CAAC,sBAAsB,CAAC,WAAW,EAAE,IAAI,EAAE,QAAQ,CAAC,EAAE,MAAM,CAAC,CAAA;IACtF,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAA;IAE7B,6FAA6F;IAC7F,KAAK,MAAM,IAAI,IAAI,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;QAC/C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAA;QAE3B,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,SAAQ;QAEjD,8DAA8D;QAC9D,KAAK,MAAM,IAAI,IAAI,OAAO,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC;YACvC,MAAM,CAAC,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAA;YAEnC,IAAI,CAAC,EAAE,CAAC;gBACP,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;YACX,CAAC;QACF,CAAC;IACF,CAAC;IAED,OAAO,GAAG,CAAA;AACX,CAAC;AAED,MAAM,YAAY,GAAG,cAAc,CAAC,kBAAkB,CAAC,CAAA;AACvD,MAAM,YAAY,GAAG,cAAc,CAAC,kBAAkB,CAAC,CAAA;AAQvD;;;;;GAKG;AACH,MAAM,UAAU,eAAe,CAAC,QAAgB;IAC/C,MAAM,OAAO,GAAG,QAAQ,CAAC,IAAI,EAAE,CAAA;IAE/B,IAAI,CAAC,OAAO;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IAE/D,MAAM,MAAM,GAAG,OAAO,CAAC,KAAK,CAAC,KAAK,CAAC,CAAA;IAEnC,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IAE/E,MAAM,IAAI,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAA;IAE9D,IAAI,MAAM,GAAkB,IAAI,CAAA;IAChC,IAAI,MAAM,GAAkB,IAAI,CAAA;IAChC,IAAI,QAAQ,GAAG,CAAC,CAAA;IAChB,IAAI,MAAM,GAAG,MAAM,CAAC,MAAM,CAAA;IAE1B,6BAA6B;IAC7B,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC,CAAC,IAAI,MAAM,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC;QAC9D,MAAM,GAAG,MAAM,CAAC,CAAC,CAAE,CAAA;QACnB,QAAQ,GAAG,CAAC,CAAA;IACb,CAAC;IAED,mFAAmF;IACnF,MAAM,IAAI,GAAG,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAC,CAAA;IACtC,MAAM,UAAU,GAAG,MAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAA;IAE/D,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,YAAY,CAAC,GAAG,CAAC,UAAU,CAAC,EAAE,CAAC;QAC5D,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;QACnD,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;SAAM,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,MAAM,GAAG,QAAQ,IAAI,CAAC,EAAE,CAAC;QAC7D,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAA;QAC5B,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;SAAM,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,MAAM,GAAG,QAAQ,IAAI,CAAC,EAAE,CAAC;QAC7D,gCAAgC;QAChC,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAA;QAC5B,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;IAED,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IAE9D,IAAI,CAAC,MAAM,EAAE,CAAC;QACb,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IACvD,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,CAAA;AAClC,CAAC"}
|
|
@@ -8,9 +8,9 @@
|
|
|
8
8
|
* Federally Qualified Health Centers (FQHCs) are HRSA-funded community health programs that
|
|
9
9
|
* self-report site addresses to the HRSA Data Warehouse. The published CSV (`data.hrsa.gov`)
|
|
10
10
|
* carries the site name, the postal-formatted street address, and the locality/region/postcode
|
|
11
|
-
* quad.
|
|
12
|
-
*
|
|
13
|
-
*
|
|
11
|
+
* quad. The source earns its place on adversarial-value-per-row (#22): every facility name is a
|
|
12
|
+
* human-typed venue string and the addresses pass through enough hands to accumulate the
|
|
13
|
+
* abbreviation drift + suite designator chaos that pure gazetteer data does not.
|
|
14
14
|
*
|
|
15
15
|
* The adapter consumes a CSV file the operator pre-downloads. The HRSA data is published as a
|
|
16
16
|
* single national CSV (~10K rows), small enough that the operator can re-fetch on every corpus
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;
|
|
1
|
+
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAOH,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAEjF;;;GAGG;AACH,eAAO,MAAM,0BAA0B,oBAAoB,CAAA;AAC3D;;;GAGG;AACH,eAAO,MAAM,+BAA+B,kBAAkB,CAAA;AAmD9D,wBAAgB,0BAA0B,IAAI,aAAa,CA8E1D;AAED;;GAEG;AACH,eAAO,MAAM,oBAAoB,eAA+B,CAAA"}
|
|
@@ -8,9 +8,9 @@
|
|
|
8
8
|
* Federally Qualified Health Centers (FQHCs) are HRSA-funded community health programs that
|
|
9
9
|
* self-report site addresses to the HRSA Data Warehouse. The published CSV (`data.hrsa.gov`)
|
|
10
10
|
* carries the site name, the postal-formatted street address, and the locality/region/postcode
|
|
11
|
-
* quad.
|
|
12
|
-
*
|
|
13
|
-
*
|
|
11
|
+
* quad. The source earns its place on adversarial-value-per-row (#22): every facility name is a
|
|
12
|
+
* human-typed venue string and the addresses pass through enough hands to accumulate the
|
|
13
|
+
* abbreviation drift + suite designator chaos that pure gazetteer data does not.
|
|
14
14
|
*
|
|
15
15
|
* The adapter consumes a CSV file the operator pre-downloads. The HRSA data is published as a
|
|
16
16
|
* single national CSV (~10K rows), small enough that the operator can re-fetch on every corpus
|
|
@@ -28,9 +28,8 @@
|
|
|
28
28
|
* License: stamped `"Public Domain"` per the HRSA Data Warehouse's federal government distribution
|
|
29
29
|
* terms.
|
|
30
30
|
*/
|
|
31
|
-
import {
|
|
32
|
-
import {
|
|
33
|
-
import { stableSourceID } from "../../adapter.js";
|
|
31
|
+
import { CSVSpliterator } from "spliterator";
|
|
32
|
+
import { splitStreetLine, stableSourceID } from "../../adapter.js";
|
|
34
33
|
import { lookupStateAbbreviation } from "../../codex/us-fips-state.js";
|
|
35
34
|
import { reconcileComponents } from "../../format.js";
|
|
36
35
|
/**
|
|
@@ -47,21 +46,11 @@ export const USGOV_HRSA_FQHC_DEFAULT_LICENSE = "Public Domain";
|
|
|
47
46
|
* Split a "123 Main St Suite 4" surface form into `(house_number, street)`. The regex tolerates one trailing letter on
|
|
48
47
|
* the number (`"123A Main St"`) and a hyphenated form (`"40-12 Bell Blvd"`); anything else falls back to street-only.
|
|
49
48
|
*
|
|
50
|
-
* Suite / Apt / Unit designators stay on `street`
|
|
51
|
-
*
|
|
52
|
-
*
|
|
53
|
-
*
|
|
49
|
+
* Suite / Apt / Unit designators stay on `street` — Mailwoman's `unit` component exists but the address-formatter does
|
|
50
|
+
* not have a clean slot for it, and HRSA addresses do not separate the suite into its own column. Leaving the surface
|
|
51
|
+
* form intact in `street` preserves the adversarial training signal (the model learns that a trailing "Suite 4" is part
|
|
52
|
+
* of the road line in this distribution).
|
|
54
53
|
*/
|
|
55
|
-
const HOUSE_NUMBER_PREFIX = /^(\d+(?:-\d+)?[A-Za-z]?)\s+(.+)$/;
|
|
56
|
-
function splitAddress(address) {
|
|
57
|
-
const trimmed = address.trim();
|
|
58
|
-
if (!trimmed)
|
|
59
|
-
return null;
|
|
60
|
-
const m = HOUSE_NUMBER_PREFIX.exec(trimmed);
|
|
61
|
-
if (m)
|
|
62
|
-
return { house_number: m[1], street: m[2].trim() };
|
|
63
|
-
return { street: trimmed };
|
|
64
|
-
}
|
|
65
54
|
/**
|
|
66
55
|
* Compose the raw envelope-style address line. Format:
|
|
67
56
|
*
|
|
@@ -84,67 +73,60 @@ export function createUsgovHrsaFqhcAdapter() {
|
|
|
84
73
|
if (opts.country && opts.country !== "US") {
|
|
85
74
|
throw new Error(`usgov-hrsa-fqhc adapter: only US supported, got country=${opts.country}`);
|
|
86
75
|
}
|
|
87
|
-
const
|
|
88
|
-
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
|
|
92
|
-
relax_column_count: true,
|
|
93
|
-
}));
|
|
76
|
+
const rows = CSVSpliterator.fromAsync(opts.inputPath, {
|
|
77
|
+
mode: "object",
|
|
78
|
+
normalizeKeys: false,
|
|
79
|
+
enableQuoteHandling: true,
|
|
80
|
+
});
|
|
94
81
|
let emitted = 0;
|
|
95
|
-
|
|
96
|
-
|
|
97
|
-
|
|
98
|
-
|
|
99
|
-
|
|
100
|
-
|
|
101
|
-
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
|
|
111
|
-
|
|
112
|
-
|
|
113
|
-
|
|
114
|
-
|
|
115
|
-
|
|
116
|
-
|
|
117
|
-
|
|
118
|
-
|
|
119
|
-
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
|
|
123
|
-
|
|
124
|
-
|
|
125
|
-
|
|
126
|
-
|
|
127
|
-
|
|
128
|
-
|
|
129
|
-
|
|
130
|
-
|
|
131
|
-
|
|
132
|
-
|
|
133
|
-
|
|
134
|
-
|
|
135
|
-
|
|
136
|
-
|
|
137
|
-
|
|
138
|
-
|
|
139
|
-
|
|
140
|
-
|
|
141
|
-
|
|
142
|
-
|
|
143
|
-
emitted++;
|
|
144
|
-
}
|
|
145
|
-
}
|
|
146
|
-
finally {
|
|
147
|
-
stream.destroy();
|
|
82
|
+
for await (const record of rows) {
|
|
83
|
+
if (opts.signal?.aborted)
|
|
84
|
+
break;
|
|
85
|
+
if (opts.limit !== undefined && emitted >= opts.limit)
|
|
86
|
+
break;
|
|
87
|
+
const venue = (record["Site Name"] ?? "").trim();
|
|
88
|
+
const split = splitStreetLine(record["Site Address"] ?? "");
|
|
89
|
+
const city = (record["Site City"] ?? "").trim();
|
|
90
|
+
const stateAbbr = (record["Site State Abbreviation"] ?? "").trim();
|
|
91
|
+
const postcode = (record["Site Postal Code"] ?? "").trim();
|
|
92
|
+
if (!venue || !split || !city || !postcode)
|
|
93
|
+
continue;
|
|
94
|
+
const state = lookupStateAbbreviation(stateAbbr);
|
|
95
|
+
if (!state)
|
|
96
|
+
continue;
|
|
97
|
+
// Insertion order matters here. `venue` first so alignment claims its span
|
|
98
|
+
// (which may contain a token like "Buffalo") before `locality` runs its
|
|
99
|
+
// search — the kryptonite case `Buffalo Health Clinic, Buffalo NY`
|
|
100
|
+
// otherwise mis-labels the venue's "Buffalo" as locality.
|
|
101
|
+
const components = {
|
|
102
|
+
venue,
|
|
103
|
+
...(split.house_number ? { house_number: split.house_number } : {}),
|
|
104
|
+
street: split.street,
|
|
105
|
+
locality: city,
|
|
106
|
+
region: state.abbreviation,
|
|
107
|
+
postcode,
|
|
108
|
+
};
|
|
109
|
+
const raw = composeRaw(venue, split.house_number, split.street, city, state.abbreviation, postcode);
|
|
110
|
+
if (!raw)
|
|
111
|
+
continue;
|
|
112
|
+
const aligned = reconcileComponents(components, raw);
|
|
113
|
+
if (!Object.keys(aligned).length)
|
|
114
|
+
continue;
|
|
115
|
+
const siteID = (record["Site ID"] ?? "").trim();
|
|
116
|
+
const sourceID = siteID
|
|
117
|
+
? `${USGOV_HRSA_FQHC_ADAPTER_ID}-${siteID}`
|
|
118
|
+
: stableSourceID(USGOV_HRSA_FQHC_ADAPTER_ID, aligned);
|
|
119
|
+
yield {
|
|
120
|
+
raw,
|
|
121
|
+
components: aligned,
|
|
122
|
+
country: "US",
|
|
123
|
+
locale: "en-US",
|
|
124
|
+
source: USGOV_HRSA_FQHC_ADAPTER_ID,
|
|
125
|
+
source_id: sourceID,
|
|
126
|
+
corpus_version: "",
|
|
127
|
+
license: USGOV_HRSA_FQHC_DEFAULT_LICENSE,
|
|
128
|
+
};
|
|
129
|
+
emitted++;
|
|
148
130
|
}
|
|
149
131
|
},
|
|
150
132
|
};
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAEH,OAAO,EAAE,
|
|
1
|
+
{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../src/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAA;AAE5C,OAAO,EAAE,eAAe,EAAE,cAAc,EAAE,MAAM,kBAAkB,CAAA;AAClE,OAAO,EAAE,uBAAuB,EAAE,MAAM,8BAA8B,CAAA;AACtE,OAAO,EAAE,mBAAmB,EAAE,MAAM,iBAAiB,CAAA;AAGrD;;;GAGG;AACH,MAAM,CAAC,MAAM,0BAA0B,GAAG,iBAAiB,CAAA;AAC3D;;;GAGG;AACH,MAAM,CAAC,MAAM,+BAA+B,GAAG,eAAe,CAAA;AAmB9D;;;;;;;;GAQG;AAEH;;;;;;;GAOG;AACH,SAAS,UAAU,CAClB,KAAa,EACb,KAAyB,EACzB,MAAc,EACd,IAAY,EACZ,KAAa,EACb,QAAgB;IAEhB,MAAM,UAAU,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IACnE,MAAM,QAAQ,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,EAAE,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;IAE7G,OAAO,CAAC,KAAK,CAAC,IAAI,EAAE,EAAE,UAAU,EAAE,QAAQ,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;AACvE,CAAC;AAED,MAAM,UAAU,0BAA0B;IACzC,OAAO;QACN,EAAE,EAAE,0BAA0B;QAC9B,cAAc,EAAE,+BAA+B;QAC/C,WAAW,EACV,yIAAyI;QAE1I,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,KAAK,CAAC,2DAA2D,IAAI,CAAC,OAAO,EAAE,CAAC,CAAA;YAC3F,CAAC;YAED,MAAM,IAAI,GAAG,cAAc,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,EAAE;gBACrD,IAAI,EAAE,QAAQ;gBACd,aAAa,EAAE,KAAK;gBACpB,mBAAmB,EAAE,IAAI;aACzB,CAAC,CAAA;YAEF,IAAI,OAAO,GAAG,CAAC,CAAA;YAEf,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,IAAkC,EAAE,CAAC;gBAC/D,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAE/B,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;oBAAE,MAAK;gBAE5D,MAAM,KAAK,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAChD,MAAM,KAAK,GAAG,eAAe,CAAC,MAAM,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAA;gBAC3D,MAAM,IAAI,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAC/C,MAAM,SAAS,GAAG,CAAC,MAAM,CAAC,yBAAyB,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAClE,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,kBAAkB,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE1D,IAAI,CAAC,KAAK,IAAI,CAAC,KAAK,IAAI,CAAC,IAAI,IAAI,CAAC,QAAQ;oBAAE,SAAQ;gBACpD,MAAM,KAAK,GAAG,uBAAuB,CAAC,SAAS,CAAC,CAAA;gBAEhD,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,2EAA2E;gBAC3E,wEAAwE;gBACxE,mEAAmE;gBACnE,0DAA0D;gBAC1D,MAAM,UAAU,GAA+B;oBAC9C,KAAK;oBACL,GAAG,CAAC,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,KAAK,CAAC,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACnE,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,QAAQ,EAAE,IAAI;oBACd,MAAM,EAAE,KAAK,CAAC,YAAY;oBAC1B,QAAQ;iBACR,CAAA;gBAED,MAAM,GAAG,GAAG,UAAU,CAAC,KAAK,EAAE,KAAK,CAAC,YAAY,EAAE,KAAK,CAAC,MAAM,EAAE,IAAI,EAAE,KAAK,CAAC,YAAY,EAAE,QAAQ,CAAC,CAAA;gBAEnG,IAAI,CAAC,GAAG;oBAAE,SAAQ;gBAElB,MAAM,OAAO,GAAG,mBAAmB,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;gBAEpD,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM;oBAAE,SAAQ;gBAE1C,MAAM,MAAM,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE/C,MAAM,QAAQ,GAAG,MAAM;oBACtB,CAAC,CAAC,GAAG,0BAA0B,IAAI,MAAM,EAAE;oBAC3C,CAAC,CAAC,cAAc,CAAC,0BAA0B,EAAE,OAAO,CAAC,CAAA;gBAEtD,MAAM;oBACL,GAAG;oBACH,UAAU,EAAE,OAAO;oBACnB,OAAO,EAAE,IAAI;oBACb,MAAM,EAAE,OAAO;oBACf,MAAM,EAAE,0BAA0B;oBAClC,SAAS,EAAE,QAAQ;oBACnB,cAAc,EAAE,EAAE;oBAClB,OAAO,EAAE,+BAA+B;iBACxC,CAAA;gBAED,OAAO,EAAE,CAAA;YACV,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,0BAA0B,EAAE,CAAA"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/usgov-imls-pls/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;
|
|
1
|
+
{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../src/adapters/usgov-imls-pls/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAOH,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,gBAAgB,CAAA;AAEjF;;;GAGG;AACH,eAAO,MAAM,yBAAyB,mBAAmB,CAAA;AACzD;;;GAGG;AACH,eAAO,MAAM,8BAA8B,kBAAkB,CAAA;AAY7D,wBAAgB,yBAAyB,IAAI,aAAa,CAqFzD;AAED;;GAEG;AACH,eAAO,MAAM,mBAAmB,eAA8B,CAAA"}
|