@larsgw/formica 0.5.0 → 0.5.2

package/.gitattributes

@@ -0,0 +1 @@
+*.ts linguist-language=TypeScript

package/.github/workflows/ci.yml

@@ -0,0 +1,27 @@
+name: Node.js CI
+
+on:
+  push:
+    branches: [ "main" ]
+  pull_request:
+    branches: [ "main" ]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    strategy:
+      matrix:
+        node-version: [18.x, 20.x]
+
+    steps:
+    - uses: actions/checkout@v3
+    - name: Use Node.js ${{ matrix.node-version }}
+      uses: actions/setup-node@v3
+      with:
+        node-version: ${{ matrix.node-version }}
+    - run: npm install
+    - run: npm run build
+    - run: npm run lint
+    - run: npm test

package/CHANGELOG.md

@@ -1,3 +1,21 @@
+## [0.5.2](https://github.com/identification-resources/formica/compare/v0.5.1...v0.5.2) (2023-11-27)
+
+
+### Bug Fixes
+
+* **resources:** fix GBIF index generation ([c88e2d2](https://github.com/identification-resources/formica/commit/c88e2d2b306b930ca0b00b978e3eb89b8adfb656))
+
+
+
+## [0.5.1](https://github.com/identification-resources/formica/compare/v0.5.0...v0.5.1) (2023-11-03)
+
+
+### Bug Fixes
+
+* **resources:** fix file listing in processor ([009b980](https://github.com/identification-resources/formica/commit/009b98091b0f597e91213ace82252facbe3b5fed))
+
+
+
 # [0.5.0](https://github.com/identification-resources/formica/compare/v0.4.3...v0.5.0) (2023-10-24)
 
 

package/README.md

@@ -3,7 +3,17 @@
 **Formica** is an SDK and a collection of tools for the data in the Library of
 Identification Resources.
 
-## Tools
+## Install
+
+Install globally:
+
+    npm install --global @larsgw/formica
+
+Run locally:
+
+    npx --package @larsgw/formica -- ...
+
+## CLI
 
   - `loir-validate-catalog [./catalog.csv ./authors.csv ./places.csv ./publishers.csv]`:
     Validate the CSV files containing metadata ([schema](https://github.com/identification-resources/catalog/blob/main/docs/tools-resources.md)).
@@ -17,3 +27,12 @@ Identification Resources.
     `--update` flag in a Git repo to update existing Darwin Core files while keeping
     the identifiers stable (in most cases).
   - `loir-resources-index [./resources]`: Create indices of the Darwin Core archives.
+
+## API
+
+```js
+const Formica = require('@larsgw/formica')
+
+// Load catalog data
+const catalog = Formica.catalog.loadData(`...`, 'catalog')
+```

package/lib/bin/process-resources-index.js

@@ -96,7 +96,7 @@ function main(args) {
                                             file = _a.sent();
                                             return [2 /*return*/, Promise.all(index_1.resources.parseTextFileHeader(file).map(function (resource, index) {
                                                     return __awaiter(this, void 0, void 0, function () {
-                                                        var amendedResource, dwcFile, _a, _header, dwc, _b, _c, _i, dwc_1, taxon, gbifId;
+                                                        var amendedResource, dwcFile, _a, header, dwc, _b, _c, gbifColumn, gbifAcceptedColumn, _i, dwc_1, taxon, gbifId;
                                                         return __generator(this, function (_d) {
                                                             switch (_d.label) {
                                                                 case 0:
@@ -108,14 +108,16 @@ function main(args) {
                                                                     _c = (_b = index_1.csv).parseCsv;
                                                                     return [4 /*yield*/, fs_1.promises.readFile(dwcFile, 'utf-8')];
                                                                 case 1:
-                                                                    _a = _c.apply(_b, [_d.sent()]), _header = _a[0], dwc = _a.slice(1);
+                                                                    _a = _c.apply(_b, [_d.sent()]), header = _a[0], dwc = _a.slice(1);
+                                                                    gbifColumn = header.indexOf('gbifTaxonID');
+                                                                    gbifAcceptedColumn = header.indexOf('gbifAcceptedTaxonID');
                                                                     for (_i = 0, dwc_1 = dwc; _i < dwc_1.length; _i++) {
                                                                         taxon = dwc_1[_i];
-                                                                        gbifId = taxon[25];
+                                                                        gbifId = taxon[gbifColumn];
                                                                         if (gbifId) {
                                                                             addTaxon(gbifIndex, gbifId, taxon);
-                                                                            if (taxon[27] !== taxon[25]) {
-                                                                                addTaxon(gbifIndex, taxon[27], taxon);
+                                                                            if (taxon[gbifAcceptedColumn] !== taxon[gbifColumn]) {
+                                                                                addTaxon(gbifIndex, taxon[gbifAcceptedColumn], taxon);
                                                                             }
                                                                         }
                                                                         amendedResource.taxonCount += 1;

package/lib/bin/process-resources.js

@@ -155,7 +155,7 @@ function listUnprocessedFiles(directory, outputDirectory) {
                     return [4 /*yield*/, fs_1.promises.readdir(outputDirectory)];
                 case 2:
                     output = new (_a.apply(Set, [void 0, _b.sent()]))();
-                    return [2 /*return*/, input.filter(function (file) { return output.has(file + '-1'); })];
+                    return [2 /*return*/, input.filter(function (file) { return !output.has(file + '-1.csv'); })];
             }
         });
     });

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@larsgw/formica",
-  "version": "0.5.0",
+  "version": "0.5.2",
   "description": "SDK and tools for data from the Library of Identification Resources",
   "main": "lib/index.js",
   "types": "lib/index.d.ts",

package/src/bin/process-resources-index.ts

@@ -54,13 +54,15 @@ async function main (args: string[]): Promise<void> {
                 return
             }
 
-            const [_header, ...dwc] = csv.parseCsv(await fs.readFile(dwcFile, 'utf-8'))
+            const [header, ...dwc] = csv.parseCsv(await fs.readFile(dwcFile, 'utf-8'))
+            const gbifColumn = header.indexOf('gbifTaxonID')
+            const gbifAcceptedColumn = header.indexOf('gbifAcceptedTaxonID')
             for (const taxon of dwc) {
-                const gbifId = taxon[25]
+                const gbifId = taxon[gbifColumn]
                 if (gbifId) {
                     addTaxon(gbifIndex, gbifId, taxon)
-                    if (taxon[27] !== taxon[25]) {
-                        addTaxon(gbifIndex, taxon[27], taxon)
+                    if (taxon[gbifAcceptedColumn] !== taxon[gbifColumn]) {
+                        addTaxon(gbifIndex, taxon[gbifAcceptedColumn], taxon)
                     }
                 }
                 amendedResource.taxonCount += 1

package/src/bin/process-resources.ts

@@ -98,7 +98,7 @@ async function listFiles (directory: string): Promise<string[]> {
 async function listUnprocessedFiles (directory: string, outputDirectory: string): Promise<string[]> {
     const input = await listFiles(directory)
     const output = new Set(await fs.readdir(outputDirectory))
-    return input.filter(file => output.has(file + '-1'))
+    return input.filter(file => !output.has(file + '-1.csv'))
 }
 
 async function listChangedFiles (directory: string): Promise<string[]> {