jats-xml 0.0.13 → 0.0.15

package/README.md

@@ -35,7 +35,7 @@ jats download https://elifesciences.org/articles/81952 article.jats
 
 Note, currently this just downloads the XML, **not** the associated files.
 
-`sumamry`: summarize the contents of the JATS, given a URL, DOI, or local file
+`summary`: summarize the contents of the JATS, given a URL, DOI, or local file
 
 ```bash
 jats summary https://elifesciences.org/articles/81952
@@ -47,6 +47,12 @@ This will provide a summary, including a list of what the JATS file contains.
 
 ![Output of `jats summary`](/images/jats-output.png)
 
+`validate`: validate local file against JATS Archive DTD schema. By default, this uses JATS 1.3.
+
+```bash
+jats validate article.jats --jats 1.2 --mathmml 2
+```
+
 ## Working in Typescript
 
 All tags are accessible as types/enums. There is also documentation from each node-type

package/dist/cjs/{download.js → cli/download.js}

@@ -14,20 +14,10 @@ var __importDefault = (this && this.__importDefault) || function (mod) {
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.downloadJatsFromUrl = exports.checkIfPubMedCentralHasJats = exports.convertPMID2PMCID = void 0;
 const doi_utils_1 = __importDefault(require("doi-utils"));
-const chalk_1 = __importDefault(require("chalk"));
 const node_fetch_1 = __importDefault(require("node-fetch"));
 const myst_cli_utils_1 = require("myst-cli-utils");
-const fair_principles_1 = require("fair-principles");
-const resolvers_1 = require("./resolvers");
-function logAboutJatsFailing(session, jatsUrls) {
-    session.log.warn('⛔️ JATS may not be Open Access 😭, you should petition your local representative 🪧');
-    session.log.info(`${chalk_1.default.green(`\nThe XML ${chalk_1.default.bold('should')} be here:\n\n${jatsUrls.join('\n')}`)}\n`);
-    const FAIR = (0, fair_principles_1.highlightFAIR)('A', { chalk: chalk_1.default });
-    session.log.info(`Some publishers aggressively block programmatic access, which isn't ${FAIR}.`);
-    session.log.debug((0, fair_principles_1.formatPrinciples)('A*', { chalk: chalk_1.default }));
-    session.log.info(`${chalk_1.default.blue('The link may work in a browser.')}\n`);
-}
-function dowloadFromUrl(session, jatsUrl, opts) {
+const resolvers_1 = require("../resolvers");
+function downloadFromUrl(session, jatsUrl, opts) {
     var _a, _b;
     return __awaiter(this, void 0, void 0, function* () {
         const toc = (0, myst_cli_utils_1.tic)();
@@ -41,7 +31,7 @@ function dowloadFromUrl(session, jatsUrl, opts) {
         if (!((contentType === null || contentType === void 0 ? void 0 : contentType.includes('application/xml')) ||
             (contentType === null || contentType === void 0 ? void 0 : contentType.includes('text/xml')) ||
             (contentType === null || contentType === void 0 ? void 0 : contentType.includes('text/plain')))) {
-            session.log.warn(`Expected content-type "application/xml" instead we got "${contentType}" for ${jatsUrl}\n${chalk_1.default.dim('Things may not work, but we are going to try our best...')}`);
+            session.log.warn(`Expected content-type "application/xml" instead we got "${contentType}" for ${jatsUrl}`);
         }
         const data = yield resp.text();
         session.log.debug(toc(`Fetched document with content-type "${contentType}" in %s`));
@@ -169,26 +159,26 @@ function downloadJatsFromUrl(session, urlOrDoi, opts = {}) {
             for (let index = 0; index < expectedUrls.length; index++) {
                 const url = expectedUrls[index];
                 try {
-                    const data = yield dowloadFromUrl(session, url, opts);
+                    const data = yield downloadFromUrl(session, url, opts);
                     if (data)
-                        return { source: url, data };
+                        return { success: true, source: url, data };
                 }
                 catch (error) {
                     session.log.debug(error.message);
                 }
             }
             // If there are expected URLs that don't work: see something, say something, etc.
-            logAboutJatsFailing(session, expectedUrls);
+            return { success: false, source: expectedUrls[0] };
         }
         if (doi_utils_1.default.validate(urlOrDoi)) {
             const jatsUrl = yield (0, resolvers_1.customResolveJatsUrlFromDoi)(session, urlOrDoi, opts);
-            const data = yield dowloadFromUrl(session, jatsUrl, opts);
-            return { source: jatsUrl, data };
+            const data = yield downloadFromUrl(session, jatsUrl, opts);
+            return { success: true, source: jatsUrl, data };
         }
         if ((0, myst_cli_utils_1.isUrl)(urlOrDoi)) {
             session.log.debug("No resolver matched, and the URL doesn't look like a DOI. We will attempt to download it directly.");
-            const data = yield dowloadFromUrl(session, urlOrDoi, opts);
-            return { source: urlOrDoi, data };
+            const data = yield downloadFromUrl(session, urlOrDoi, opts);
+            return { success: true, source: urlOrDoi, data };
         }
         throw new Error(`Could not find ${urlOrDoi} locally, and it doesn't look like a URL or DOI`);
     });

package/dist/cjs/cli/index.js

@@ -7,8 +7,10 @@ Object.defineProperty(exports, "__esModule", { value: true });
 const commander_1 = __importDefault(require("commander"));
 const version_1 = __importDefault(require("../version"));
 const parse_1 = require("./parse");
+const validate_1 = require("./validate");
 const program = new commander_1.default.Command();
 (0, parse_1.addDownloadCLI)(program);
+(0, validate_1.addValidateCLI)(program);
 program.version(`v${version_1.default}`, '-v, --version', 'Print the current version of jats-xml');
 program.option('-d, --debug', 'Log out any errors to the console.');
 program.parse(process.argv);

package/dist/cjs/cli/parse.js

@@ -19,12 +19,13 @@ const path_1 = require("path");
 const myst_cli_utils_1 = require("myst-cli-utils");
 const doi_utils_1 = __importDefault(require("doi-utils"));
 const chalk_1 = __importDefault(require("chalk"));
+const fair_principles_1 = require("fair-principles");
 const session_1 = require("../session");
 const types_1 = require("../types");
 const jats_1 = require("../jats");
 const myst_common_1 = require("myst-common");
 const unist_util_select_1 = require("unist-util-select");
-const download_1 = require("../download");
+const download_1 = require("./download");
 const resolvers_1 = require("../resolvers");
 const utils_1 = require("../utils");
 function hasValidExtension(output) {
@@ -41,11 +42,23 @@ function downloadAndSaveJats(session, urlOrDoi, output, opts = { resolvers: reso
         if (!hasValidExtension(output)) {
             session.log.warn(`The extension ${(0, path_1.extname)(output)} is not a valid extension for JATS, try using ".xml" or ".jats"`);
         }
-        const { data } = yield (0, download_1.downloadJatsFromUrl)(session, urlOrDoi, opts);
+        const { success, data, source } = yield (0, download_1.downloadJatsFromUrl)(session, urlOrDoi, opts);
+        if (!success || !data) {
+            logAboutJatsFailing(session, [source]);
+            process.exit(1);
+        }
         (0, myst_cli_utils_1.writeFileToFolder)(output, data);
         return data;
     });
 }
+function logAboutJatsFailing(session, jatsUrls) {
+    session.log.warn('⛔️ JATS may not be Open Access 😭, you should petition your local representative 🪧');
+    session.log.info(`${chalk_1.default.green(`\nThe XML ${chalk_1.default.bold('should')} be here:\n\n${jatsUrls.join('\n')}`)}\n`);
+    const FAIR = (0, fair_principles_1.highlightFAIR)('A', { chalk: chalk_1.default });
+    session.log.info(`Some publishers aggressively block programmatic access, which isn't ${FAIR}.`);
+    session.log.debug((0, fair_principles_1.formatPrinciples)('A*', { chalk: chalk_1.default }));
+    session.log.info(`\n${chalk_1.default.blue('The link may work in a browser.')}\n`);
+}
 function parseJats(session, file, opts = { resolvers: resolvers_1.DEFAULT_RESOLVERS }) {
     return __awaiter(this, void 0, void 0, function* () {
         const toc = (0, myst_cli_utils_1.tic)();
@@ -54,7 +67,11 @@ function parseJats(session, file, opts = { resolvers: resolvers_1.DEFAULT_RESOLV
             const data = fs_1.default.readFileSync(file).toString();
             return new jats_1.Jats(data, { log: session.log });
         }
-        const { source, data } = yield (0, download_1.downloadJatsFromUrl)(session, file, opts);
+        const { success, source, data } = yield (0, download_1.downloadJatsFromUrl)(session, file, opts);
+        if (!success || !data) {
+            logAboutJatsFailing(session, [source]);
+            process.exit(1);
+        }
         const jats = new jats_1.Jats(data, { source, log: session.log });
         session.log.debug(toc(`Downloaded and parsed JATS file in %s`));
         return jats;
@@ -195,7 +212,7 @@ function makeDownloadCLI(program) {
     const command = new commander_1.Command('download')
         .description('Parse a JATS file and provide a summary')
         .argument('<url>', 'The JATS url or a DOI')
-        .argument('<output>', 'The JATS url or a DOI')
+        .argument('<output>', 'The JATS output file')
         .action((0, myst_cli_utils_1.clirun)(downloadAndSaveJats, { program, getSession: session_1.getSession }));
     return command;
 }

package/dist/cjs/cli/validate.js

@@ -0,0 +1,22 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.addValidateCLI = void 0;
+const commander_1 = require("commander");
+const myst_cli_utils_1 = require("myst-cli-utils");
+const session_1 = require("../session");
+const validate_1 = require("../validate");
+function makeValidateCLI(program) {
+    const command = new commander_1.Command('validate')
+        .description('Fetch JATS DTD schema file from nih.gov ftp server')
+        .argument('<file>', 'JATS file to validate')
+        .addOption(new commander_1.Option('--jats <version>', 'JATS version, must be 1.2 or later').default('1.3'))
+        .addOption(new commander_1.Option('--mathml <version>', 'MathML version, 2 or 3').default('3'))
+        .addOption(new commander_1.Option('--oasis', 'Use OASIS table model').default(false))
+        .addOption(new commander_1.Option('--directory <value>', 'Directory to save DTD file'))
+        .action((0, myst_cli_utils_1.clirun)(validate_1.validateJatsAgainstDtdWrapper, { program, getSession: session_1.getSession }));
+    return command;
+}
+function addValidateCLI(program) {
+    program.addCommand(makeValidateCLI(program));
+}
+exports.addValidateCLI = addValidateCLI;

package/dist/cjs/index.js

@@ -17,11 +17,10 @@ var __importDefault = (this && this.__importDefault) || function (mod) {
     return (mod && mod.__esModule) ? mod : { "default": mod };
 };
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.downloadJatsFromUrl = exports.Jats = exports.version = void 0;
+exports.Jats = exports.version = void 0;
 var version_1 = require("./version");
 Object.defineProperty(exports, "version", { enumerable: true, get: function () { return __importDefault(version_1).default; } });
 var jats_1 = require("./jats");
 Object.defineProperty(exports, "Jats", { enumerable: true, get: function () { return jats_1.Jats; } });
-var download_1 = require("./download");
-Object.defineProperty(exports, "downloadJatsFromUrl", { enumerable: true, get: function () { return download_1.downloadJatsFromUrl; } });
 __exportStar(require("./types"), exports);
+__exportStar(require("./validate"), exports);

package/dist/cjs/validate/dtd.js

@@ -0,0 +1,170 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || function (mod) {
+    if (mod && mod.__esModule) return mod;
+    var result = {};
+    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
+    __setModuleDefault(result, mod);
+    return result;
+};
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.validateJatsAgainstDtdWrapper = exports.validateJatsAgainstDtd = void 0;
+const fs_1 = __importStar(require("fs"));
+const path_1 = __importDefault(require("path"));
+const node_fetch_1 = __importDefault(require("node-fetch"));
+const unzipper_1 = __importDefault(require("unzipper"));
+const which_1 = require("which");
+const myst_cli_utils_1 = require("myst-cli-utils");
+const chalk_1 = __importDefault(require("chalk"));
+const JATS_VERSIONS = ['1.2', '1.2d1', '1.2d2', '1.3', '1.3d1', '1.3d2'];
+const DEFAULT_JATS_VERSION = '1.3';
+const MATHML_VERSIONS = ['2', '3'];
+const DEFAULT_MATHML_VERSION = '3';
+function validateOptions(opts) {
+    var _a;
+    let jats;
+    if (!opts.jats) {
+        jats = DEFAULT_JATS_VERSION;
+    }
+    else if (!JATS_VERSIONS.includes(opts.jats)) {
+        throw new Error(`Invalid JATS version "${opts.jats}" - must be one of [${JATS_VERSIONS.join(', ')}]`);
+    }
+    else {
+        jats = opts.jats;
+    }
+    let mathml;
+    if (!opts.mathml) {
+        mathml = DEFAULT_MATHML_VERSION;
+    }
+    else if (!MATHML_VERSIONS.includes(opts.mathml)) {
+        throw new Error(`Invalid MathML version "${opts.mathml}" - must be one of [${MATHML_VERSIONS.join(', ')}]`);
+    }
+    else {
+        mathml = opts.mathml;
+    }
+    const out = {
+        jats,
+        mathml,
+        oasis: !!opts.oasis,
+        directory: (_a = opts.directory) !== null && _a !== void 0 ? _a : defaultDirectory(),
+    };
+    return out;
+}
+function dtdFolder(opts) {
+    const version = opts.jats.replace('.', '-');
+    const oasis = opts.oasis ? '-OASIS' : '';
+    const mathml = `MathML${opts.mathml}`;
+    return `JATS-Archiving-${version}${oasis}-${mathml}-DTD`;
+}
+function dtdZipFile(opts) {
+    return `${dtdFolder(opts)}.zip`;
+}
+function localDtdZipFile(opts) {
+    return path_1.default.join(opts.directory, dtdZipFile(opts));
+}
+function dtdFile(opts) {
+    const version = opts.jats.startsWith('1.2') ? '1' : opts.jats.replace('.', '-');
+    const article = opts.oasis ? 'archive-oasis-article' : 'archivearticle';
+    const mathml = opts.mathml === '3' ? '-mathml3' : '';
+    return `JATS-${article}${version}${mathml}.dtd`;
+}
+function localDtdFile(opts) {
+    return path_1.default.join(opts.directory, dtdFolder(opts), dtdFile(opts));
+}
+function ftpUrl(opts) {
+    return `https://ftp.ncbi.nih.gov/pub/jats/archiving/${opts.jats}/${dtdZipFile(opts)}`;
+}
+function defaultDirectory() {
+    return path_1.default.join(__dirname, 'static');
+}
+function dtdDownload(session, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (!fs_1.default.existsSync(opts.directory)) {
+            fs_1.default.mkdirSync(opts.directory, { recursive: true });
+        }
+        session.log.info(`🌎 Downloading: ${ftpUrl(opts)}`);
+        session.log.debug(`Saving to ${localDtdZipFile(opts)}`);
+        const resp = yield (0, node_fetch_1.default)(ftpUrl(opts));
+        (0, myst_cli_utils_1.writeFileToFolder)(localDtdZipFile(opts), yield resp.buffer());
+    });
+}
+function ensureDtdZipExists(session, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (!fs_1.default.existsSync(path_1.default.join(opts.directory, dtdZipFile(opts)))) {
+            yield dtdDownload(session, opts);
+        }
+    });
+}
+function ensureDtdExists(session, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (!fs_1.default.existsSync(localDtdFile(opts))) {
+            yield ensureDtdZipExists(session, opts);
+            const zipFile = localDtdZipFile(opts);
+            session.log.info(`🤐 Unzipping template on disk ${zipFile}`);
+            yield (0, fs_1.createReadStream)(zipFile)
+                .pipe(unzipper_1.default.Extract({ path: opts.directory }))
+                .promise();
+        }
+        session.log.debug(`Validating against ${localDtdFile(opts)}`);
+    });
+}
+function isXmllintAvailable() {
+    return (0, which_1.sync)('xmllint', { nothrow: true });
+}
+function validateJatsAgainstDtd(session, file, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (!isXmllintAvailable()) {
+            session.log.error(`JATS validation against DTD requires xmllint\n\n${chalk_1.default.dim('To install:\n  mac:    brew install xmlstarlet\n  debian: apt install libxml2-utils')}`);
+            return;
+        }
+        const validatedOpts = validateOptions(opts !== null && opts !== void 0 ? opts : {});
+        yield ensureDtdExists(session, validatedOpts);
+        try {
+            yield (0, myst_cli_utils_1.makeExecutable)(`xmllint --dtdvalid ${localDtdFile(validatedOpts)} --noout ${file}`, session.log)();
+        }
+        catch (_a) {
+            return false;
+        }
+        return true;
+    });
+}
+exports.validateJatsAgainstDtd = validateJatsAgainstDtd;
+function validateJatsAgainstDtdWrapper(session, file, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const success = yield validateJatsAgainstDtd(session, file, opts);
+        if (success) {
+            session.log.info(chalk_1.default.greenBright('JATS validation passed!'));
+        }
+        else {
+            throw new Error('JATS validation failed.');
+        }
+    });
+}
+exports.validateJatsAgainstDtdWrapper = validateJatsAgainstDtdWrapper;

package/dist/cjs/validate/index.js

@@ -0,0 +1,17 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __exportStar = (this && this.__exportStar) || function(m, exports) {
+    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+__exportStar(require("./dtd"), exports);

package/dist/cjs/version.js

@@ -1,4 +1,4 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-const version = '0.0.13';
+const version = '0.0.15';
 exports.default = version;

package/dist/esm/{download.js → cli/download.js}

@@ -8,20 +8,10 @@ var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, ge
     });
 };
 import doi from 'doi-utils';
-import chalk from 'chalk';
 import fetch from 'node-fetch';
 import { isUrl, tic } from 'myst-cli-utils';
-import { formatPrinciples, highlightFAIR } from 'fair-principles';
-import { customResolveJatsUrlFromDoi } from './resolvers';
-function logAboutJatsFailing(session, jatsUrls) {
-    session.log.warn('⛔️ JATS may not be Open Access 😭, you should petition your local representative 🪧');
-    session.log.info(`${chalk.green(`\nThe XML ${chalk.bold('should')} be here:\n\n${jatsUrls.join('\n')}`)}\n`);
-    const FAIR = highlightFAIR('A', { chalk });
-    session.log.info(`Some publishers aggressively block programmatic access, which isn't ${FAIR}.`);
-    session.log.debug(formatPrinciples('A*', { chalk }));
-    session.log.info(`${chalk.blue('The link may work in a browser.')}\n`);
-}
-function dowloadFromUrl(session, jatsUrl, opts) {
+import { customResolveJatsUrlFromDoi } from '../resolvers';
+function downloadFromUrl(session, jatsUrl, opts) {
     var _a, _b;
     return __awaiter(this, void 0, void 0, function* () {
         const toc = tic();
@@ -35,7 +25,7 @@ function dowloadFromUrl(session, jatsUrl, opts) {
         if (!((contentType === null || contentType === void 0 ? void 0 : contentType.includes('application/xml')) ||
             (contentType === null || contentType === void 0 ? void 0 : contentType.includes('text/xml')) ||
             (contentType === null || contentType === void 0 ? void 0 : contentType.includes('text/plain')))) {
-            session.log.warn(`Expected content-type "application/xml" instead we got "${contentType}" for ${jatsUrl}\n${chalk.dim('Things may not work, but we are going to try our best...')}`);
+            session.log.warn(`Expected content-type "application/xml" instead we got "${contentType}" for ${jatsUrl}`);
         }
         const data = yield resp.text();
         session.log.debug(toc(`Fetched document with content-type "${contentType}" in %s`));
@@ -161,26 +151,26 @@ export function downloadJatsFromUrl(session, urlOrDoi, opts = {}) {
             for (let index = 0; index < expectedUrls.length; index++) {
                 const url = expectedUrls[index];
                 try {
-                    const data = yield dowloadFromUrl(session, url, opts);
+                    const data = yield downloadFromUrl(session, url, opts);
                     if (data)
-                        return { source: url, data };
+                        return { success: true, source: url, data };
                 }
                 catch (error) {
                     session.log.debug(error.message);
                 }
             }
             // If there are expected URLs that don't work: see something, say something, etc.
-            logAboutJatsFailing(session, expectedUrls);
+            return { success: false, source: expectedUrls[0] };
         }
         if (doi.validate(urlOrDoi)) {
             const jatsUrl = yield customResolveJatsUrlFromDoi(session, urlOrDoi, opts);
-            const data = yield dowloadFromUrl(session, jatsUrl, opts);
-            return { source: jatsUrl, data };
+            const data = yield downloadFromUrl(session, jatsUrl, opts);
+            return { success: true, source: jatsUrl, data };
         }
         if (isUrl(urlOrDoi)) {
             session.log.debug("No resolver matched, and the URL doesn't look like a DOI. We will attempt to download it directly.");
-            const data = yield dowloadFromUrl(session, urlOrDoi, opts);
-            return { source: urlOrDoi, data };
+            const data = yield downloadFromUrl(session, urlOrDoi, opts);
+            return { success: true, source: urlOrDoi, data };
         }
         throw new Error(`Could not find ${urlOrDoi} locally, and it doesn't look like a URL or DOI`);
     });

package/dist/esm/cli/index.js

@@ -2,8 +2,10 @@
 import commander from 'commander';
 import version from '../version';
 import { addDownloadCLI } from './parse';
+import { addValidateCLI } from './validate';
 const program = new commander.Command();
 addDownloadCLI(program);
+addValidateCLI(program);
 program.version(`v${version}`, '-v, --version', 'Print the current version of jats-xml');
 program.option('-d, --debug', 'Log out any errors to the console.');
 program.parse(process.argv);

package/dist/esm/cli/parse.js

@@ -13,12 +13,13 @@ import { extname } from 'path';
 import { clirun, isUrl, tic, writeFileToFolder } from 'myst-cli-utils';
 import doi from 'doi-utils';
 import chalk from 'chalk';
+import { formatPrinciples, highlightFAIR } from 'fair-principles';
 import { getSession } from '../session';
 import { Tags } from '../types';
 import { Jats } from '../jats';
 import { toText } from 'myst-common';
 import { select, selectAll } from 'unist-util-select';
-import { downloadJatsFromUrl } from '../download';
+import { downloadJatsFromUrl } from './download';
 import { DEFAULT_RESOLVERS } from '../resolvers';
 import { findArticleId, formatDate, toDate } from '../utils';
 function hasValidExtension(output) {
@@ -35,11 +36,23 @@ function downloadAndSaveJats(session, urlOrDoi, output, opts = { resolvers: DEFA
         if (!hasValidExtension(output)) {
             session.log.warn(`The extension ${extname(output)} is not a valid extension for JATS, try using ".xml" or ".jats"`);
         }
-        const { data } = yield downloadJatsFromUrl(session, urlOrDoi, opts);
+        const { success, data, source } = yield downloadJatsFromUrl(session, urlOrDoi, opts);
+        if (!success || !data) {
+            logAboutJatsFailing(session, [source]);
+            process.exit(1);
+        }
         writeFileToFolder(output, data);
         return data;
     });
 }
+function logAboutJatsFailing(session, jatsUrls) {
+    session.log.warn('⛔️ JATS may not be Open Access 😭, you should petition your local representative 🪧');
+    session.log.info(`${chalk.green(`\nThe XML ${chalk.bold('should')} be here:\n\n${jatsUrls.join('\n')}`)}\n`);
+    const FAIR = highlightFAIR('A', { chalk });
+    session.log.info(`Some publishers aggressively block programmatic access, which isn't ${FAIR}.`);
+    session.log.debug(formatPrinciples('A*', { chalk }));
+    session.log.info(`\n${chalk.blue('The link may work in a browser.')}\n`);
+}
 function parseJats(session, file, opts = { resolvers: DEFAULT_RESOLVERS }) {
     return __awaiter(this, void 0, void 0, function* () {
         const toc = tic();
@@ -48,7 +61,11 @@ function parseJats(session, file, opts = { resolvers: DEFAULT_RESOLVERS }) {
             const data = fs.readFileSync(file).toString();
             return new Jats(data, { log: session.log });
         }
-        const { source, data } = yield downloadJatsFromUrl(session, file, opts);
+        const { success, source, data } = yield downloadJatsFromUrl(session, file, opts);
+        if (!success || !data) {
+            logAboutJatsFailing(session, [source]);
+            process.exit(1);
+        }
         const jats = new Jats(data, { source, log: session.log });
         session.log.debug(toc(`Downloaded and parsed JATS file in %s`));
         return jats;
@@ -189,7 +206,7 @@ function makeDownloadCLI(program) {
     const command = new Command('download')
         .description('Parse a JATS file and provide a summary')
         .argument('<url>', 'The JATS url or a DOI')
-        .argument('<output>', 'The JATS url or a DOI')
+        .argument('<output>', 'The JATS output file')
         .action(clirun(downloadAndSaveJats, { program, getSession }));
     return command;
 }

package/dist/esm/cli/validate.js

@@ -0,0 +1,18 @@
+import { Command, Option } from 'commander';
+import { clirun } from 'myst-cli-utils';
+import { getSession } from '../session';
+import { validateJatsAgainstDtdWrapper } from '../validate';
+function makeValidateCLI(program) {
+    const command = new Command('validate')
+        .description('Fetch JATS DTD schema file from nih.gov ftp server')
+        .argument('<file>', 'JATS file to validate')
+        .addOption(new Option('--jats <version>', 'JATS version, must be 1.2 or later').default('1.3'))
+        .addOption(new Option('--mathml <version>', 'MathML version, 2 or 3').default('3'))
+        .addOption(new Option('--oasis', 'Use OASIS table model').default(false))
+        .addOption(new Option('--directory <value>', 'Directory to save DTD file'))
+        .action(clirun(validateJatsAgainstDtdWrapper, { program, getSession }));
+    return command;
+}
+export function addValidateCLI(program) {
+    program.addCommand(makeValidateCLI(program));
+}

package/dist/esm/index.js

@@ -1,4 +1,4 @@
 export { default as version } from './version';
 export { Jats } from './jats';
-export { downloadJatsFromUrl } from './download';
 export * from './types';
+export * from './validate';