@seo-partner/adapters 0.0.0-stage → 1.0.3

package/src/parsers/dtl-parser.ts

@@ -0,0 +1,166 @@
+import type {
+  DiscoveredRoute,
+  ExtractedRouteData,
+  HeadingItem,
+  ImageItem,
+  JsonLdItem,
+  LinkItem,
+} from '@seo-partner/core';
+
+function lineNumberAtOffset(content: string, offset: number): number {
+  let line = 1;
+  for (let i = 0; i < offset && i < content.length; i += 1) {
+    if (content.charCodeAt(i) === 10) line += 1;
+  }
+  return line;
+}
+
+function getAttr(attrsRaw: string, attrName: string): string | null {
+  const re = new RegExp(`\\b${attrName}\\s*=\\s*["']([^"']*)["']`, 'i');
+  const m = re.exec(attrsRaw);
+  return m ? (m[1] ?? null) : null;
+}
+
+export interface DtlParseResult {
+  readonly routeData: ExtractedRouteData;
+  readonly extendsTemplate?: string;
+}
+
+export function parseDjangoTemplate(
+  route: DiscoveredRoute,
+  fileContent: string,
+): DtlParseResult {
+  let title: { value: string; line: number } | null = null;
+  let description: { value: string; line: number } | null = null;
+  let canonical: { value: string; line: number } | null = null;
+  let robots: { value: string; line: number } | null = null;
+  let htmlLang: { value: string; line: number } | null | undefined = undefined;
+  let isDynamic = false;
+  let extendsTemplate: string | undefined = undefined;
+
+  const openGraph: Record<string, string> = {};
+  const hreflang: Array<{ lang: string; url: string }> = [];
+  const headings: HeadingItem[] = [];
+  const images: ImageItem[] = [];
+  const links: LinkItem[] = [];
+  const jsonLdItems: JsonLdItem[] = [];
+
+  // 1. Check for {% extends 'base.html' %}
+  const extendsMatch = /\{%\s*extends\s+['"]([^'"]+)['"]\s*%\}/i.exec(fileContent);
+  if (extendsMatch && extendsMatch[1]) {
+    extendsTemplate = extendsMatch[1].replace(/\\/g, '/');
+  }
+
+  // 2. Check for {% block title %}...{% endblock %}
+  const blockTitleMatch = /\{%\s*block\s+title\s*%\}([\s\S]*?)\{%\s*endblock\s*(?:title)?\s*%\}/i.exec(fileContent);
+  if (blockTitleMatch && blockTitleMatch[1]) {
+    const rawTitle = blockTitleMatch[1].trim();
+    const line = lineNumberAtOffset(fileContent, blockTitleMatch.index);
+    if (/\{\{[^}]+\}\}/.test(rawTitle)) {
+      title = { value: '[Valeur dynamique]', line };
+      isDynamic = true;
+    } else {
+      title = { value: rawTitle.replace(/\s+/g, ' '), line };
+    }
+  }
+
+  // 3. Check for direct <title>...</title>
+  if (!title) {
+    const directTitleMatch = /<title[^>]*>([\s\S]*?)<\/title>/i.exec(fileContent);
+    if (directTitleMatch && directTitleMatch[1]) {
+      const rawTitle = directTitleMatch[1].trim();
+      const line = lineNumberAtOffset(fileContent, directTitleMatch.index);
+      if (/\{\{[^}]+\}\}/.test(rawTitle) || /\{%/.test(rawTitle)) {
+        title = { value: '[Valeur dynamique]', line };
+        isDynamic = true;
+      } else {
+        title = { value: rawTitle.replace(/\s+/g, ' '), line };
+      }
+    }
+  }
+
+  // 4. Check for <html lang="...">
+  const htmlTagMatch = /<html\b([^>]*?)>/i.exec(fileContent);
+  if (htmlTagMatch) {
+    const langVal = getAttr(htmlTagMatch[1] ?? '', 'lang');
+    if (langVal) {
+      htmlLang = {
+        value: langVal.trim(),
+        line: lineNumberAtOffset(fileContent, htmlTagMatch.index),
+      };
+    }
+  }
+
+  // 5. Meta tags: <meta name="description" ...>
+  const metaRegex = /<meta\b([^>]*?)\/?>/gi;
+  let metaMatch: RegExpExecArray | null;
+  while ((metaMatch = metaRegex.exec(fileContent)) !== null) {
+    const attrs = metaMatch[1] ?? '';
+    const line = lineNumberAtOffset(fileContent, metaMatch.index);
+    const nameVal = getAttr(attrs, 'name')?.toLowerCase();
+    const propVal = getAttr(attrs, 'property')?.toLowerCase();
+    const contentVal = getAttr(attrs, 'content');
+    if (nameVal === 'description' && contentVal !== null && !description) {
+      description = { value: contentVal, line };
+    } else if (nameVal === 'robots' && contentVal !== null && !robots) {
+      robots = { value: contentVal, line };
+    } else if (propVal?.startsWith('og:') && contentVal !== null) {
+      openGraph[propVal.slice(3)] = contentVal;
+    }
+  }
+
+  // 6. Headings: <h1> to <h6>
+  const headingRegex = /<h([1-6])\b[^>]*>([\s\S]*?)<\/h\1>/gi;
+  let hMatch: RegExpExecArray | null;
+  while ((hMatch = headingRegex.exec(fileContent)) !== null) {
+    const level = Number(hMatch[1]) as 1 | 2 | 3 | 4 | 5 | 6;
+    const text = (hMatch[2] ?? '')
+      .replace(/<[^>]+>/g, '')
+      .replace(/\{[{%][\s\S]*?[}%]\}/g, '')
+      .replace(/\s+/g, ' ')
+      .trim();
+    headings.push({
+      level,
+      text,
+      line: lineNumberAtOffset(fileContent, hMatch.index),
+    });
+  }
+
+  // 7. Links: <a href="...">
+  const aRegex = /<a\b([^>]*?)>([\s\S]*?)<\/a>/gi;
+  let aMatch: RegExpExecArray | null;
+  while ((aMatch = aRegex.exec(fileContent)) !== null) {
+    const attrs = aMatch[1] ?? '';
+    const href = getAttr(attrs, 'href');
+    if (href) {
+      const anchorText = (aMatch[2] ?? '').replace(/<[^>]+>/g, '').trim();
+      links.push({
+        toUrl: href,
+        anchorText,
+        kind: href.startsWith('http://') || href.startsWith('https://') ? 'external' : 'internal',
+        rel: getAttr(attrs, 'rel'),
+        line: lineNumberAtOffset(fileContent, aMatch.index),
+      });
+    }
+  }
+
+  return {
+    routeData: {
+      route,
+      title,
+      description,
+      canonical,
+      robots,
+      htmlLang,
+      headings,
+      openGraph,
+      hreflang,
+      images,
+      links,
+      jsonLdItems,
+      isDynamic,
+    },
+    extendsTemplate,
+  };
+}
+

package/src/parsers/html-parser.ts

@@ -0,0 +1,214 @@
+import type {
+  DiscoveredRoute,
+  ExtractedRouteData,
+  HeadingItem,
+  ImageItem,
+  JsonLdItem,
+  LinkItem,
+} from '@seo-partner/core';
+
+function lineNumberAtOffset(content: string, offset: number): number {
+  let line = 1;
+  for (let i = 0; i < offset && i < content.length; i += 1) {
+    if (content.charCodeAt(i) === 10) line += 1;
+  }
+  return line;
+}
+
+function getAttr(attrsRaw: string, attrName: string): string | null {
+  const re = new RegExp(`\\b${attrName}\\s*=\\s*["']([^"']*)["']`, 'i');
+  const m = re.exec(attrsRaw);
+  return m ? (m[1] ?? null) : null;
+}
+
+export function parseStaticHtmlOrMarkdown(
+  route: DiscoveredRoute,
+  fileContent: string,
+): ExtractedRouteData {
+  let title: { value: string; line: number } | null = null;
+  let description: { value: string; line: number } | null = null;
+  let canonical: { value: string; line: number } | null = null;
+  let robots: { value: string; line: number } | null = null;
+  let htmlLang: { value: string; line: number } | null | undefined = undefined;
+  const openGraph: Record<string, string> = {};
+  const hreflang: Array<{ lang: string; url: string }> = [];
+  const headings: HeadingItem[] = [];
+  const images: ImageItem[] = [];
+  const links: LinkItem[] = [];
+  const jsonLdItems: JsonLdItem[] = [];
+
+  if (route.routerKind === 'markdown') {
+    const lines = fileContent.split(/\r?\n/);
+    for (let i = 0; i < lines.length; i += 1) {
+      const lineText = lines[i] ?? '';
+      const lineNum = i + 1;
+      const titleFm = /^title:\s*["']?(.+?)["']?$/i.exec(lineText.trim());
+      if (titleFm && !title) {
+        title = { value: (titleFm[1] ?? '').trim(), line: lineNum };
+      }
+      const descFm = /^description:\s*["']?(.+?)["']?$/i.exec(lineText.trim());
+      if (descFm && !description) {
+        description = { value: (descFm[1] ?? '').trim(), line: lineNum };
+      }
+      const hMatch = /^(#{1,6})\s+(.+)$/.exec(lineText.trim());
+      if (hMatch) {
+        const level = (hMatch[1]?.length ?? 1) as 1 | 2 | 3 | 4 | 5 | 6;
+        headings.push({ level, text: (hMatch[2] ?? '').trim(), line: lineNum });
+      }
+    }
+    return {
+      route,
+      title,
+      description,
+      canonical,
+      robots,
+      headings,
+      openGraph,
+      hreflang,
+      images,
+      links,
+      jsonLdItems,
+    };
+  }
+
+  const htmlTagMatch = /<html\b([^>]*?)>/i.exec(fileContent);
+  if (htmlTagMatch) {
+    const langVal = getAttr(htmlTagMatch[1] ?? '', 'lang') ?? '';
+    htmlLang = {
+      value: langVal.trim(),
+      line: lineNumberAtOffset(fileContent, htmlTagMatch.index),
+    };
+  }
+
+  const titleMatch = /<title[^>]*>([\s\S]*?)<\/title>/i.exec(fileContent);
+  if (titleMatch) {
+    title = {
+      value: (titleMatch[1] ?? '').replace(/\s+/g, ' ').trim(),
+      line: lineNumberAtOffset(fileContent, titleMatch.index),
+    };
+  }
+
+  const metaRegex = /<meta\b([^>]*?)\/?>/gi;
+  let metaMatch: RegExpExecArray | null;
+  while ((metaMatch = metaRegex.exec(fileContent)) !== null) {
+    const attrs = metaMatch[1] ?? '';
+    const line = lineNumberAtOffset(fileContent, metaMatch.index);
+    const nameVal = getAttr(attrs, 'name')?.toLowerCase();
+    const propVal = getAttr(attrs, 'property')?.toLowerCase();
+    const contentVal = getAttr(attrs, 'content');
+    if (nameVal === 'description' && contentVal !== null && !description) {
+      description = { value: contentVal, line };
+    } else if (nameVal === 'robots' && contentVal !== null && !robots) {
+      robots = { value: contentVal, line };
+    } else if (propVal?.startsWith('og:') && contentVal !== null) {
+      openGraph[propVal.slice(3)] = contentVal;
+    }
+  }
+
+  const linkTagRegex = /<link\b([^>]*?)\/?>/gi;
+  let linkTagMatch: RegExpExecArray | null;
+  while ((linkTagMatch = linkTagRegex.exec(fileContent)) !== null) {
+    const attrs = linkTagMatch[1] ?? '';
+    const line = lineNumberAtOffset(fileContent, linkTagMatch.index);
+    const relVal = getAttr(attrs, 'rel')?.toLowerCase();
+    const hrefVal = getAttr(attrs, 'href');
+    if (relVal === 'canonical' && hrefVal !== null && !canonical) {
+      canonical = { value: hrefVal, line };
+    } else if (relVal === 'alternate' && hrefVal !== null) {
+      const hl = getAttr(attrs, 'hreflang');
+      if (hl) hreflang.push({ lang: hl, url: hrefVal });
+    }
+  }
+
+  const headingRegex = /<h([1-6])\b[^>]*>([\s\S]*?)<\/h\1>/gi;
+  let hMatch: RegExpExecArray | null;
+  while ((hMatch = headingRegex.exec(fileContent)) !== null) {
+    const level = Number(hMatch[1]) as 1 | 2 | 3 | 4 | 5 | 6;
+    const text = (hMatch[2] ?? '').replace(/<[^>]+>/g, '').replace(/\s+/g, ' ').trim();
+    headings.push({
+      level,
+      text,
+      line: lineNumberAtOffset(fileContent, hMatch.index),
+    });
+  }
+
+  const imgRegex = /<img\b([^>]*?)\/?>/gi;
+  let imgMatch: RegExpExecArray | null;
+  while ((imgMatch = imgRegex.exec(fileContent)) !== null) {
+    const attrs = imgMatch[1] ?? '';
+    const srcVal = getAttr(attrs, 'src') ?? '';
+    const altVal = getAttr(attrs, 'alt');
+    images.push({
+      src: srcVal,
+      alt: altVal,
+      hasAlt: altVal !== null && altVal.trim().length > 0,
+      width: getAttr(attrs, 'width'),
+      height: getAttr(attrs, 'height'),
+      line: lineNumberAtOffset(fileContent, imgMatch.index),
+    });
+  }
+
+  const anchorRegex = /<a\b([^>]*?)>([\s\S]*?)<\/a>/gi;
+  let aMatch: RegExpExecArray | null;
+  while ((aMatch = anchorRegex.exec(fileContent)) !== null) {
+    const attrs = aMatch[1] ?? '';
+    const hrefVal = getAttr(attrs, 'href');
+    if (hrefVal) {
+      links.push({
+        toUrl: hrefVal,
+        anchorText: (aMatch[2] ?? '').replace(/<[^>]+>/g, '').replace(/\s+/g, ' ').trim(),
+        kind: /^https?:\/\//i.test(hrefVal) ? 'external' : 'internal',
+        rel: getAttr(attrs, 'rel'),
+        line: lineNumberAtOffset(fileContent, aMatch.index),
+      });
+    }
+  }
+
+  const scriptRegex = /<script\b[^>]*type=["']application\/ld\+json["'][^>]*>([\s\S]*?)<\/script>/gi;
+  let sMatch: RegExpExecArray | null;
+  while ((sMatch = scriptRegex.exec(fileContent)) !== null) {
+    const rawJson = (sMatch[1] ?? '').trim();
+    const line = lineNumberAtOffset(fileContent, sMatch.index);
+    try {
+      const parsed = JSON.parse(rawJson) as Record<string, unknown>;
+      const contextVal = parsed['@context'];
+      const typeVal = parsed['@type'];
+      const isValid =
+        typeof contextVal === 'string' &&
+        contextVal.includes('schema.org') &&
+        typeof typeVal === 'string' &&
+        typeVal.trim().length > 0;
+      jsonLdItems.push({
+        schemaType: typeof typeVal === 'string' ? typeVal : null,
+        rawJson,
+        isValid,
+        validationError: isValid ? null : 'Missing @context (schema.org) or @type',
+        line,
+      });
+    } catch (err) {
+      jsonLdItems.push({
+        schemaType: null,
+        rawJson,
+        isValid: false,
+        validationError: err instanceof Error ? err.message : 'Invalid JSON',
+        line,
+      });
+    }
+  }
+
+  return {
+    route,
+    title,
+    description,
+    canonical,
+    robots,
+    ...(htmlLang !== undefined ? { htmlLang } : {}),
+    headings,
+    openGraph,
+    hreflang,
+    images,
+    links,
+    jsonLdItems,
+  };
+}
+

package/src/parsers/layout-resolver.ts

@@ -0,0 +1,112 @@
+import path from 'node:path';
+import type { ExtractedRouteData } from '@seo-partner/core';
+
+export interface ExtractedLayoutData {
+  readonly filePath: string;
+  readonly htmlLang?: { readonly value: string; readonly line: number } | null;
+  readonly title?: { readonly value: string; readonly line: number } | null;
+  readonly description?: { readonly value: string; readonly line: number } | null;
+  readonly canonical?: { readonly value: string; readonly line: number } | null;
+  readonly robots?: { readonly value: string; readonly line: number } | null;
+}
+
+export class LayoutResolver {
+  private readonly layoutsByPath = new Map<string, ExtractedLayoutData>();
+
+  public registerLayout(layout: ExtractedLayoutData): void {
+    const normalized = layout.filePath.replace(/\\/g, '/');
+    this.layoutsByPath.set(normalized, layout);
+  }
+
+  public getLayout(filePath: string): ExtractedLayoutData | undefined {
+    const normalized = filePath.replace(/\\/g, '/');
+    return this.layoutsByPath.get(normalized);
+  }
+
+  /**
+   * Applies layout inheritance to an extracted route if the route lacks metadata
+   * (such as htmlLang or fallback title).
+   */
+  public applyInheritance(
+    routeData: ExtractedRouteData,
+    parentLayoutPaths: ReadonlyArray<string>,
+  ): ExtractedRouteData {
+    let htmlLang = routeData.htmlLang;
+    let title = routeData.title;
+    let description = routeData.description;
+    let canonical = routeData.canonical;
+    let robots = routeData.robots;
+
+    // Traverse from nearest layout to root layout
+    for (const layoutPath of parentLayoutPaths) {
+      const layout = this.getLayout(layoutPath);
+      if (!layout) continue;
+
+      if (!htmlLang && layout.htmlLang) {
+        htmlLang = layout.htmlLang;
+      }
+      if (!title && layout.title) {
+        title = layout.title;
+      }
+      if (!description && layout.description) {
+        description = layout.description;
+      }
+      if (!canonical && layout.canonical) {
+        canonical = layout.canonical;
+      }
+      if (!robots && layout.robots) {
+        robots = layout.robots;
+      }
+    }
+
+    return {
+      ...routeData,
+      htmlLang,
+      title,
+      description,
+      canonical,
+      robots,
+    };
+  }
+
+  /**
+   * Finds candidate SvelteKit layout paths (+layout.svelte) starting from the page directory
+   * up to the routes root directory.
+   */
+  public findSvelteKitLayoutPaths(pageFilePath: string): ReadonlyArray<string> {
+    const normalized = pageFilePath.replace(/\\/g, '/');
+    const parts = normalized.split('/');
+    const routesIdx = parts.lastIndexOf('routes');
+    if (routesIdx === -1) return [];
+
+    const candidates: string[] = [];
+    const dirParts = parts.slice(0, -1);
+
+    // From current dir up to routes directory
+    for (let i = dirParts.length; i >= routesIdx; i--) {
+      const dir = dirParts.slice(0, i).join('/');
+      candidates.push(`${dir}/+layout.svelte`);
+    }
+
+    return candidates;
+  }
+
+  /**
+   * Resolves Django base template paths (e.g. "base.html" or "layouts/base.html").
+   */
+  public findDjangoExtendsPath(
+    templateContent: string,
+    allTemplatePaths: ReadonlyArray<string>,
+  ): string | undefined {
+    const extendsMatch = templateContent.match(/\{%\s*extends\s+['"]([^'"]+)['"]\s*%\}/i);
+    if (!extendsMatch || !extendsMatch[1]) return undefined;
+
+    const baseName = extendsMatch[1].replace(/\\/g, '/');
+    // Match against known templates
+    return allTemplatePaths.find((p) => {
+      const norm = p.replace(/\\/g, '/');
+      return norm.endsWith(baseName) || norm === baseName;
+    });
+  }
+}
+