Manual scope & how to read¶
Where the content comes from¶
Everything under Official User Manual is extracted directly from the shipped EAR package:
cpm-5.3.36.549/
└─ gsw_ef.war
└─ WEB-INF/lib/tgk-user-manual-5.3.36-6.jar
└─ META-INF/resources/static/man/en/ ← the online manual, source of this site
| Property | Value |
|---|---|
| Product release | CCH Tagetik 5.3.36.549 (SP36) |
| Chapters | all 13 |
| TOC entries | 2,519 |
| Pages extracted | 2,145 |
| Figures | 2,083 (deduplicated by content hash, ~37 MB) |
| Search index | ~7,000 entries |
| Languages shipped | en / fr / it / jp — this site builds en and zh |
The TOC entry count is higher than the page count because the official table of contents references some pages from several places.
Chapters¶
| # | Chapter | Pages |
|---|---|---|
| 1 | Platform Management | 852 |
| 2 | Extended Planning | 402 |
| 3 | Reporting & Data Visualization | 380 |
| 4 | Data Entry | 253 |
| 5 | Financial Close & Consolidation | 228 |
| 6 | Data Integration | 205 |
| 7 | Intelligent Disclosure | 156 |
| 8 | Getting Started | 38 |
| 9 | Homepage / What's New / This manual / Using the Search Feature / Glossary | 1 each |
Counts are TOC entries; a few entries share a page, and 15 entries have no page at all (see below).
What the extraction preserves¶
- section hierarchy (heading levels normalised so the page title is the only
h1) - cross-references — in-manual links are rewritten to site paths
- figures — copied out of the jar and served locally, no dependency on the app server
- per-page
source:front matter pointing back at the original.htmpath
Reading conventions¶
- Left navigation mirrors the official table of contents, including its nesting. It is pruned to the current branch — with 2,500 entries a fully expanded tree in every page would be unusable.
- Tabs at the top give one entry point per chapter.
- Search is indexed for the whole manual (~7,000 entries).
- Language selector switches between Chinese and English for the same topic, where a translation exists.
15 pages missing upstream¶
These pages are listed in the official table of contents but have no file in the English language pack — they exist only in other languages. This is an upstream gap, not an extraction failure:
Content/Allocazioni e Closing/… 14 pages (allocation & closing)
Content/Diagnostico/Esecuz/DIA_FinEseProcMD_r.htm 1 page
They are removed from the navigation so nothing dead-ends. The full list is written to
build/missing_upstream.json every time the extraction runs.
Known imperfections¶
These come from the upstream documentation generator (MadCap Flare), not from the extraction:
- Some figures are referenced with a wrong path upstream and are broken on the online server too. Wherever the file could still be located inside the package, this site resolves it — a few pictures here actually render better than online.
- Anchor fragments (in-page
#...links) use MadCap's internal anchor names, which do not match the generated heading slugs. Such links land on the correct page but not at the exact section. - A handful of cross-references point at the 15 pages listed above.
Extending the site¶
See the maintenance guide (Chinese) — rebuilding for a new service pack means changing one path and re-running four scripts.