Verified against baseline fixtures in InGestX Pro v1.0.0. Support does not imply complete formatting or layout reconstruction.
| Source | What you can extract | Current limits |
|---|---|---|
| Native page text and local English OCR for pages with little native text. | Advanced layout, table reconstruction, and mixed-page image OCR are not included. | |
| DOCX | Body headings, paragraphs, lists, tables, and media inventory. | Full formatting, headers/footers, and image binary export are not reconstructed. |
| PPTX | Slide text, notes, tables, and image inventory. | Complex layout and image binary export are not reconstructed. |
| XLS/XLSX | Sheets, cells, formulas, stored values, and cell references. | Formulas are not recalculated; chart and structured table extraction are not included. |
| CSV | Rows, headers, quoted values, and line references. | Delimiter inference is heuristic; normalized output has size limits. |
| XML | Hierarchy, attributes, and normalized node paths. | DTDs and entities are prohibited. |
| TXT | Text blocks, line ranges, and supported text encodings. | Language is not inferred. |
| HTML | Headings, paragraphs, lists, tables, links, and code. | JavaScript is not rendered. |
| Markdown | Headings, lists, code fences, tables, and front matter. | Inline links and images remain inline content. |
| RTF | Text, Unicode, paragraphs, and table delimiters. | Full formatting and embedded assets are not reconstructed. |
| EPUB | Metadata and chapters in reading order. | Layout and binary assets are not exported. |
| JSON/JSONL | Raw objects, normalized records, and line provenance. | Malformed records report errors; large results are memory bounded. |
| Web URLs | Static HTML and bounded same-origin website crawls. | Authenticated and JavaScript-rendered sites are not supported. |
| Code Repositories | Local files, folder paths, hashes, code line ranges, and heuristic symbol hints. | Remote cloning, authentication, full AST analysis, and Git history are not included. |
| Audio (WAV, MP3, FLAC, AIFF, AAC/M4A, OGG, OPUS, WMA) | Audio ingestion, format conversion, and metadata capture. | WMA and some legacy codecs depend on available system/media decoders. |
| Images (JPG, PNG, WEBP, TIFF, BMP, GIF, HEIC/HEIF, AVIF, ICO) | Image ingestion, format conversion, and metadata capture. | HEIC/HEIF support depends on available system codecs. |
| Video (MP4, MOV, MKV, AVI, WEBM, M4V) | Video ingestion, probing, and conversion, including still image + audio to video creation. | Playback and rendering rely on standard compatible graphics drivers. |
Normalized results are capped at 100,000 blocks / 128 MiB. Very large inputs can fail explicitly. Review warnings before relying on output.