@opencraw/core 0.1.2 → 0.1.4

package/dist/xml-document/xml-parser.client.d.ts

@@ -0,0 +1,49 @@
+/** A parsed document, as the XPath engine sees it. */
+export type XmlNode = Node;
+/**
+ * Parses XML strictly but safely: a malformed document is an error naming
+ * the URL and the line; entities a DOCTYPE declares are never expanded (no
+ * "billion laughs") and nothing external is ever fetched (no XXE), they stay
+ * as they are.
+ *
+ * @param xml - The document text.
+ * @param where - The URL or id, for messages.
+ * @returns The document.
+ * @throws Error when the text is not well-formed XML.
+ */
+export declare function parseXml(xml: string, where: string): Document;
+/**
+ * HTML as an XML document, for XPath on fetched pages: parsed the way a
+ * browser parses it (a forgiving HTML5 parser that inserts `<tbody>` and
+ * closes what the page left open), without namespaces, so `//table/tbody/tr`
+ * finds what it finds on the live page.
+ *
+ * @param html - A page or a fragment.
+ * @returns The document.
+ */
+export declare function htmlAsXml(html: string): Document;
+/**
+ * A copy of a document with every namespace dropped: elements and attributes
+ * keep their local names, `xmlns` declarations go. What `ignoreNamespaces`
+ * queries run on: `//entry/title` instead of `//atom:entry/atom:title`.
+ *
+ * @param document - A parsed document.
+ * @returns The copy.
+ */
+export declare function withoutNamespaces(document: Document): Document;
+/**
+ * A node as markup: the element with its children (`outer`), or its children only.
+ *
+ * @param node - A node.
+ * @param outer - Whether to include the node itself.
+ * @returns The markup.
+ */
+export declare function serializeXml(node: Node, outer: boolean): string;
+/**
+ * The text of an XML body, gunzipped first when it is gzip (a `sitemap.xml.gz`).
+ *
+ * @param bytes - The body.
+ * @returns The bytes to decode.
+ */
+export declare function gunzipIfNeeded(bytes: Uint8Array): Uint8Array;
+//# sourceMappingURL=xml-parser.client.d.ts.map

package/dist/xml-document/xpath.algorithm.d.ts

@@ -0,0 +1,30 @@
+import type { Take } from '../selection/index.js';
+/** An XPath result: a node, or what a function such as `count()` or `string()` gives. */
+export type XpathValue = Node | string | number | boolean;
+export interface XpathOptions {
+    /** Prefix to URI, for queries on namespaced documents (`atom: http://www.w3.org/2005/Atom`). */
+    namespaces?: Record<string, string>;
+    /** Drop the document's namespaces first, so plain names match (`//entry/title`). */
+    ignoreNamespaces?: boolean;
+}
+/**
+ * Runs an XPath 1.0 query on a parsed document. Prefixes the root element
+ * declares are known without being listed; a default namespace has no
+ * prefix, so name one in `namespaces` or use `ignoreNamespaces`.
+ *
+ * @param document - The document.
+ * @param expression - The query.
+ * @param options - Namespaces.
+ * @returns Every node selected, or the one value a function returned.
+ * @throws Error naming the query when it does not parse or uses an unknown prefix.
+ */
+export declare function selectXpath(document: Document, expression: string, options?: XpathOptions): XpathValue[];
+/**
+ * The value of an XPath result.
+ *
+ * @param value - A node or a function result.
+ * @param take - `text` (whitespace collapsed), `html` (the node's inner markup), `json` (its outer markup), `value` (its text as is), `attr:<name>`.
+ * @returns The value; `undefined` for a missing attribute.
+ */
+export declare function takeFromXml(value: XpathValue, take: Take): unknown;
+//# sourceMappingURL=xpath.algorithm.d.ts.map

package/dist/yaml-document/index.d.ts

@@ -0,0 +1,3 @@
+export { readYaml } from './read-yaml.client.js';
+export type { YamlRead, YamlScalars } from './read-yaml.client.js';
+//# sourceMappingURL=index.d.ts.map

package/dist/yaml-document/read-yaml.client.d.ts

@@ -0,0 +1,25 @@
+/** How YAML scalars are read: typed as YAML 1.2 types them, or as the text written. */
+export type YamlScalars = 'typed' | 'text';
+/** A YAML text, read. */
+export interface YamlRead {
+    /** The document's value; several documents (`---`) give an array of their values. */
+    data: unknown;
+    documents: number;
+    /** What the parser noticed but read anyway: an unknown tag (`!!js/function`) is kept as its plain value. */
+    warnings: string[];
+}
+/**
+ * Parses YAML with the `yaml` package, imported on first use. The version is
+ * pinned to YAML 1.2 (core schema) whatever the document declares: under a
+ * `%YAML 1.1` directive, `NO` would read as `false` and `0123` as octal `83`.
+ * Merge keys (`<<: *base`) are applied, duplicate keys are an error, aliases
+ * are capped, and custom tags never build values: nothing in the text runs.
+ *
+ * @param text - The YAML.
+ * @param source - Where it came from, for messages.
+ * @param scalars - `typed` (default), or `text` to keep every scalar as written (`0123` stays `"0123"`).
+ * @returns The data, the number of documents and the warnings.
+ * @throws Error naming the source, with the line and column, for YAML that does not parse.
+ */
+export declare function readYaml(text: string, source: string, scalars?: YamlScalars): Promise<YamlRead>;
+//# sourceMappingURL=read-yaml.client.d.ts.map

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@opencraw/core",
-  "version": "0.1.2",
+  "version": "0.1.4",
   "type": "module",
   "main": "./dist/index.esm.js",
   "module": "./dist/index.esm.js",
@@ -20,13 +20,15 @@
     "!**/*.js.map"
   ],
   "dependencies": {
-    "@opencraw/office-reader": "^0.0.2",
+    "@opencraw/office-reader": "^0.0.3",
+    "@xmldom/xmldom": "^0.9.12",
     "cheerio": "^1.2.0",
     "domhandler": "^6.0.1",
     "jsonpath-plus": "^10.4.0",
     "marked": "^18.0.14",
     "pdfjs-dist": "^6.3.289",
     "playwright": "^1.63.0",
+    "xpath": "^0.0.34",
     "yaml": "^2.9.1",
     "zod": "^4.6.5"
   },