diff options
Diffstat (limited to 'src/data/yaml.js')
| -rw-r--r-- | src/data/yaml.js | 309 |
1 files changed, 298 insertions, 11 deletions
diff --git a/src/data/yaml.js b/src/data/yaml.js index a5614ea6..79602faa 100644 --- a/src/data/yaml.js +++ b/src/data/yaml.js @@ -11,8 +11,10 @@ import {colors, ENABLE_COLOR, logInfo, logWarn} from '#cli'; import {sortByName} from '#sort'; import Thing from '#thing'; import thingConstructors from '#things'; +import {matchContentEntries, multipleLyricsDetectionRegex} from '#wiki-data'; import { + aggregateThrows, annotateErrorWithFile, decorateErrorWithIndex, decorateErrorWithAnnotation, @@ -88,6 +90,10 @@ function makeProcessDocument(thingConstructor, { // A or B. // invalidFieldCombinations = [], + + // Bouncing function used to process subdocuments: this is a function which + // in turn calls the appropriate *result of* makeProcessDocument. + processDocument: bouncer, }) { if (!thingConstructor) { throw new Error(`Missing Thing class`); @@ -97,6 +103,10 @@ function makeProcessDocument(thingConstructor, { throw new Error(`Expected fields to be provided`); } + if (!bouncer) { + throw new Error(`Missing processDocument bouncer`); + } + const knownFields = Object.keys(fieldSpecs); const ignoredFields = @@ -144,9 +154,12 @@ function makeProcessDocument(thingConstructor, { : `document`); const aggregate = openAggregate({ + ...aggregateThrows(ProcessDocumentError), message: `Errors processing ${constructorPart}` + namePart, }); + const thing = Reflect.construct(thingConstructor, []); + const documentEntries = Object.entries(document) .filter(([field]) => !ignoredFields.includes(field)); @@ -194,13 +207,50 @@ function makeProcessDocument(thingConstructor, { const fieldValues = {}; + const subdocSymbol = Symbol('subdoc'); + const subdocLayouts = {}; + + const isSubdocToken = value => + typeof value === 'object' && + value !== null && + Object.hasOwn(value, subdocSymbol); + + const transformUtilities = { + ...thingConstructors, + + subdoc(documentType, data, { + bindInto = null, + provide = null, + } = {}) { + if (!documentType) + throw new Error(`Expected document type, got ${typeAppearance(documentType)}`); + if (!data) + throw new Error(`Expected data, got ${typeAppearance(data)}`); + if (typeof data !== 'object' || data === null) + throw new Error(`Expected data to be an object, got ${typeAppearance(data)}`); + if (typeof bindInto !== 'string' && bindInto !== null) + throw new Error(`Expected bindInto to be a string, got ${typeAppearance(bindInto)}`); + if (typeof provide !== 'object' && provide !== null) + throw new Error(`Expected provide to be an object, got ${typeAppearance(provide)}`); + + return { + [subdocSymbol]: { + documentType, + data, + bindInto, + provide, + }, + }; + }, + }; + for (const [field, documentValue] of documentEntries) { if (skippedFields.has(field)) continue; // This variable would like to certify itself as "not into capitalism". let propertyValue = (fieldSpecs[field].transform - ? fieldSpecs[field].transform(documentValue) + ? fieldSpecs[field].transform(documentValue, transformUtilities) : documentValue); // Completely blank items in a YAML list are read as null. @@ -223,10 +273,99 @@ function makeProcessDocument(thingConstructor, { } } + if (isSubdocToken(propertyValue)) { + subdocLayouts[field] = propertyValue[subdocSymbol]; + continue; + } + + if (Array.isArray(propertyValue) && propertyValue.every(isSubdocToken)) { + subdocLayouts[field] = + propertyValue + .map(token => token[subdocSymbol]); + continue; + } + fieldValues[field] = propertyValue; } - const thing = Reflect.construct(thingConstructor, []); + const subdocErrors = []; + + const followSubdocSetup = setup => { + let error = null; + + let subthing; + try { + const result = bouncer(setup.data, setup.documentType); + subthing = result.thing; + result.aggregate.close(); + } catch (caughtError) { + error = caughtError; + } + + if (subthing) { + if (setup.bindInto) { + subthing[setup.bindInto] = thing; + } + + if (setup.provide) { + Object.assign(subthing, setup.provide); + } + } + + return {error, subthing}; + }; + + for (const [field, layout] of Object.entries(subdocLayouts)) { + if (Array.isArray(layout)) { + const subthings = []; + let anySucceeded = false; + let anyFailed = false; + + for (const [index, setup] of layout.entries()) { + const {subthing, error} = followSubdocSetup(setup); + if (error) { + subdocErrors.push(new SubdocError( + {field, index}, + setup, + {cause: error})); + } + + if (subthing) { + subthings.push(subthing); + anySucceeded = true; + } else { + anyFailed = true; + } + } + + if (anySucceeded) { + fieldValues[field] = subthings; + } else if (anyFailed) { + skippedFields.add(field); + } + } else { + const setup = layout; + const {subthing, error} = followSubdocSetup(setup); + + if (error) { + subdocErrors.push(new SubdocError( + {field}, + setup, + {cause: error})); + } + + if (subthing) { + fieldValues[field] = subthing; + } else { + skippedFields.add(field); + } + } + } + + if (!empty(subdocErrors)) { + aggregate.push(new SubdocAggregateError( + subdocErrors, thingConstructor)); + } const fieldValueErrors = []; @@ -260,6 +399,8 @@ function makeProcessDocument(thingConstructor, { }); } +export class ProcessDocumentError extends AggregateError {} + export class UnknownFieldsError extends Error { constructor(fields) { super(`Unknown fields ignored: ${fields.map(field => colors.red(field)).join(', ')}`); @@ -347,12 +488,46 @@ export class SkippedFieldsSummaryError extends Error { : `${entries.length} fields`); super( - colors.bright(colors.yellow(`Altogether, skipped ${numFieldsText}:\n`)) + + colors.bright(colors.yellow(`Altogether, skipped ${numFieldsText}:`)) + '\n' + lines.join('\n') + '\n' + colors.bright(colors.yellow(`See above errors for details.`))); } } +export class SubdocError extends Error { + constructor({field, index = null}, setup, options) { + const fieldText = + (index === null + ? colors.green(`"${field}"`) + : colors.yellow(`#${index + 1}`) + ' in ' + + colors.green(`"${field}"`)); + + const constructorText = + setup.documentType.name; + + if (options.cause instanceof ProcessDocumentError) { + options.cause[Symbol.for('hsmusic.aggregate.translucent')] = true; + } + + super( + `Errors processing ${constructorText} for ${fieldText} field`, + options); + } +} + +export class SubdocAggregateError extends AggregateError { + [Symbol.for('hsmusic.aggregate.translucent')] = true; + + constructor(errors, thingConstructor) { + const constructorText = + colors.green(thingConstructor.name); + + super( + errors, + `Errors processing subdocuments for ${constructorText}`); + } +} + export function parseDate(date) { return new Date(date); } @@ -615,6 +790,108 @@ export function parseAnnotatedReferences(entries, { }); } +export function parseArtwork({ + single = false, + dimensionsFromThingProperty = null, + fileExtensionFromThingProperty = null, + dateFromThingProperty = null, + artistContribsFromThingProperty = null, + artistContribsArtistProperty = null, + artTagsFromThingProperty = null, + referencedArtworksFromThingProperty = null, +}) { + const provide = { + dimensionsFromThingProperty, + fileExtensionFromThingProperty, + dateFromThingProperty, + artistContribsFromThingProperty, + artistContribsArtistProperty, + artTagsFromThingProperty, + referencedArtworksFromThingProperty, + }; + + const parseSingleEntry = (entry, {subdoc, Artwork}) => + subdoc(Artwork, entry, {bindInto: 'thing', provide}); + + const transform = (value, ...args) => + (Array.isArray(value) + ? value.map(entry => parseSingleEntry(entry, ...args)) + : single + ? parseSingleEntry(value, ...args) + : [parseSingleEntry(value, ...args)]); + + transform.provide = provide; + + return transform; +} + +export function parseContentEntries(thingClass, sourceText, {subdoc}) { + const map = matchEntry => ({ + 'Artists': + matchEntry.artistReferences + .split(',') + .map(ref => ref.trim()), + + 'Artist Text': + matchEntry.artistDisplayText, + + 'Annotation': + matchEntry.annotation, + + 'Date': + matchEntry.date, + + 'Second Date': + matchEntry.secondDate, + + 'Date Kind': + matchEntry.dateKind, + + 'Access Date': + matchEntry.accessDate, + + 'Access Kind': + matchEntry.accessKind, + + 'Body': + matchEntry.body, + }); + + const documents = + matchContentEntries(sourceText) + .map(matchEntry => + withEntries( + map(matchEntry), + entries => entries + .filter(([key, value]) => + value !== undefined && + value !== null))); + + const subdocs = + documents.map(document => + subdoc(thingClass, document, {bindInto: 'thing'})); + + return subdocs; +} + +export function parseCommentary(sourceText, {subdoc, CommentaryEntry}) { + return parseContentEntries(CommentaryEntry, sourceText, {subdoc}); +} + +export function parseCreditingSources(sourceText, {subdoc, CreditingSourcesEntry}) { + return parseContentEntries(CreditingSourcesEntry, sourceText, {subdoc}); +} + +export function parseLyrics(sourceText, {subdoc, LyricsEntry}) { + if (!multipleLyricsDetectionRegex.test(sourceText)) { + const document = {'Body': sourceText}; + + return [subdoc(LyricsEntry, document, {bindInto: 'thing'})]; + } + + return parseContentEntries(LyricsEntry, sourceText, {subdoc}); +} + // documentModes: Symbols indicating sets of behavior for loading and processing // data files. export const documentModes = { @@ -899,7 +1176,7 @@ export function processThingsFromDataStep(documents, dataStep) { throw new Error(`Class "${thingClass.name}" doesn't specify Thing.yamlDocumentSpec`); } - fn = makeProcessDocument(thingClass, spec); + fn = makeProcessDocument(thingClass, {...spec, processDocument}); submap.set(thingClass, fn); } @@ -1280,8 +1557,7 @@ export function linkWikiDataArrays(wikiData, {bindFind, bindReverse}) { // link if the 'find' or 'reverse' properties will be implicitly linked ['albumData', [ - 'albumData', - 'trackData', + 'artworkData', 'wikiInfo', ]], @@ -1289,6 +1565,12 @@ export function linkWikiDataArrays(wikiData, {bindFind, bindReverse}) { ['artistData', [/* find, reverse */]], + ['artworkData', ['artworkData']], + + ['commentaryData', [/* find */]], + + ['creditingSourceData', [/* find */]], + ['flashData', [ 'wikiInfo', ]], @@ -1303,8 +1585,10 @@ export function linkWikiDataArrays(wikiData, {bindFind, bindReverse}) { ['homepageLayout.sections.rows', [/* find */]], + ['lyricsData', [/* find */]], + ['trackData', [ - 'albumData', + 'artworkData', 'trackData', 'wikiInfo', ]], @@ -1571,14 +1855,16 @@ export function flattenThingLayoutToDocumentOrder(layout) { } export function* splitDocumentsInYAMLSourceText(sourceText) { - const dividerRegex = /^-{3,}\n?/gm; + // Not multiline! + const dividerRegex = /(?:\r\n|\n|^)-{3,}(?:\r\n|\n|$)/g; + let previousDivider = ''; while (true) { const {lastIndex} = dividerRegex; const match = dividerRegex.exec(sourceText); if (match) { - const nextDivider = match[0].trim(); + const nextDivider = match[0]; yield { previousDivider, @@ -1589,11 +1875,12 @@ export function* splitDocumentsInYAMLSourceText(sourceText) { previousDivider = nextDivider; } else { const nextDivider = ''; + const lineBreak = previousDivider.match(/\r?\n/)?.[0] ?? ''; yield { previousDivider, nextDivider, - text: sourceText.slice(lastIndex).replace(/(?<!\n)$/, '\n'), + text: sourceText.slice(lastIndex).replace(/(?<!\n)$/, lineBreak), }; return; @@ -1619,7 +1906,7 @@ export function recombineDocumentsIntoYAMLSourceText(documents) { for (const document of documents) { if (sourceText) { - sourceText += divider + '\n'; + sourceText += divider; } sourceText += document.text; |