Processor to get list of internal links/ids and theirs title ($processor$: page-headings)

[Arpita-Jaiswal]

Status: In Progress

Owner: @Heulitig
Started On: 8th Sept 2022

Discord Thread

@Heulitig

TODO:

• need to add an option to disable auto-numbering of headings whenever required
• need to do more code refactoring and testing

13 October

• created these PR's for the changes done after 10th October
  Auto numbering of headings fastn#448
  minor adjustments #560

12 October

• collected headings from components with auto-generated ids
• auto-numbering of headings

11 October

• fixed an issue to fetch heading once per invoked component,
  no duplicate headings with different numbering from the same component

10th October

• finalized and refactored the progress till now
• Merged PR's -
  Page heading processor fastn#439,
  adjustments based on page headings processor #549

6th October

• adjusted the UI to look and behave similar to Wikipedia internal links
• fixed displaying invalid toc-item when invoking heading-list

4th October

• adjusted fix for wrong heading number
• refactored code to avoid pyramid of doom (successive if nesting)

3rd October

• removed functionality of extracting text (headings) from markdown components.
• converted linear heading list -> tree nesting for sub-headings ..
• included region check on container before extracting out their headings.
• included depth numbering for sub-headings, sub-sub-headings .. etc.

30th September

• page-headings processor partially finished
• currently returns flat list of headings (no sub-headings based on tree)

29th September

• Created a PR to add page-headings processor - Page heading processor fastn#439
• Resolved title for container components with defined regions
• Resolved undefined processor variable issue while interpreting

28th September

• Working on lazily evaluating page-headings processor to avoid double interpretation overhead.
• Resolved title and link for kernel markdown components.

@AviralVerma13

PR

22nd September:

• $processor$: internal-link -> trying to use fetch_id_from_doc to get document id, and then update global ids of the document using global_ids function from config.rs

18th September:

• $processor$: internal-link -> it can now return a list of ToC Items in the current document with their correct ids and anchor links. (but without nesting)

14th September:

• $processor$: internal-link -> return a list of ToC Items in the current document with their anchor links. (without nesting)

10th September:

• fpm: Processor to get list of internal links/ids and theirs title #487

30th Aug 2022

---

We need to have a processor that fetches all the internal-links/ids of a page with their title. For instance:

;; in fpm.ftd

-- record page-headings:
caption title:
string url:

;; in a.ftd
-- import: fpm

-- fpm.toc-item list il:
$processor$: page-headings

-- ftd.text: Title1
id: t1

-- ftd.text: Title2
id: t2

In the above case, the page-headings processor would give output something like this

[
  {
    "title": "Title1",
    "url": "a/#t1"
  },
  {
    "title": "Title2",
    "url": "a/#t2"
  }
]

This processor can help us to create a toc for page-headings.
Examples:

• Table of contents in chrome developer
• Contents in Wikipedia

[amitu]

We def need this one!

@Arpita-Jaiswal Can we reuse the toc-item instead of creating a new record?

Also the table of content should be nested, it should show each hearing and sub heading in a tree format.

[Arpita-Jaiswal]

Sounds perfect... We definitely can reuse toc-item.
So new code would look something like this:

;; in a.ftd
-- import: fpm

-- fpm.toc-item list il:
$processor$: internal-links

-- ftd.text: Title1
id: t1

-- ftd.text: Title2
id: t2

[amitu]

And for nesting we will need heading level / region stuff. Should this bit be included in current implementation? Without this we would be different from every other intra page nav tree implementation (as all else are tree and we would not be).

[Arpita-Jaiswal]

-- h0: Title1
id: t1

-- h1: Title2
id: t2


-- ftd.column h0:
caption title:
open: true
append-at: foo

--- ftd.text: $title

--- ftd.column:
id: foo



-- ftd.column h1:
caption title:
open: true
append-at: foo

--- ftd.text: $title

--- ftd.column:
id: foo

Output by $processor$: page-headings

[
  {
    "title": "Title1",
    "url": "a/#t1",
    "children": [
      {
        "title": "Title2",
        "url": "a/#t2"
      }
    ]
  }
]

[Heulitig]

Discussion (30th September)

• Had a discussion with arpita and abrar
• The page-headings processor is partially done.
• Currently it is flat heading list (No tree based nesting) and can be used to fetch headings from the document.

[amitu]

@Heulitig can you not update the original description?

[amitu]

We do status update by updating status section in description. Comments is for comment on design / feature etc.