<feed xmlns='http://www.w3.org/2005/Atom'>
<title>sisudoc-ocda-lib, branch claude-main-prepare-fix</title>
<subtitle>SiSU Spine: tools</subtitle>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/'/>
<entry>
<title>release 2.1.0: versions, a nix-gleam build, README</title>
<updated>2026-10-22T15:03:42+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-10-22T15:03:32+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=5ebd680c676b0f8cc8d59b69c2b5c8ac4e30967a'/>
<id>5ebd680c676b0f8cc8d59b69c2b5c8ac4e30967a</id>
<content type='text'>
The library's version follows the format's: 2.1.0 in dub.json,
gleam.toml and both flakes (it was 0.1.0). A release is numbered
major.minor.patch, major.minor being the format and shared with the
specification and the writer, the patch the library's own.

The Gleam build now uses nix-gleam's builder, by Arnar Gauti Ingason
(MIT), kept in gleam/nix-gleam/ with its licence and a note of where
and at which commit it came from. It fetches each Hex dependency on
its own with the checksum gleam already wrote into manifest.toml, so
there is no hash in flake.nix to keep up to date. It replaces a fixed
output derivation that ran gleam deps download, which had two faults:
its hash had to be changed by hand with every dependency, and its
output was not the same bytes from one run to the next, gleam writing
packages.toml in hash-map order. A machine that had built once reused
its output and never noticed; a fresh one would have failed the hash
at random. The programs it builds give the same output as before, for
every document-language of the markup document samples.

gleam/flake.lock: flake.nix names the nixpkgs channel url, but the
lock still recorded the github pin it replaced, so every build
rewrote it. The lock as nix rewrote it is the one built and checked;
it moves nixpkgs forward to 26.11pre1088030.

README:

- the two earlier readers, sisudoc-ssp-reader and
  sisudoc-ocda-db-reader, are named as what they now are: published
  frozen, the record of readers written from the specification
  alone, their work done by ocda-read;
- the documents the checks run over are the markup document samples,
  named with their repository and the spine flags that build the
  artefacts; the claims say "every document-language" rather than a
  count, which changes as the samples do; the same in gleam/README.md,
  which also describes the nix-gleam build and names the
  specification's documents as .md, as they are;
- "one implementation per language", where it said "one reader";
- a short section on building, a link to the specification, the
  format as 2.1, and the licence.

claude-org/: the code, the READMEs and the build files are kept as
files from here on, and org is kept for setup. config-lib.org there
tangles only .gitignore and the COPYRIGHT files, and the version and
header file there has its two local git+file blocks taken out. Both
are proposed replacements for their org/ counterparts; org/d-lib.org
and org/gleam-lib.org, which tangled the code, retire with them.

.gitignore admits a LICENSE below the root, for nix-gleam's.

Checked: D and Gleam build; with the writer at the same point, over
the markup document samples: the four dumps (D and Gleam, .ssp and
.ocda.db) identical for every document-language, ocda-read the same
from every form and in both languages, and D == Gleam in all five
formats.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The library's version follows the format's: 2.1.0 in dub.json,
gleam.toml and both flakes (it was 0.1.0). A release is numbered
major.minor.patch, major.minor being the format and shared with the
specification and the writer, the patch the library's own.

The Gleam build now uses nix-gleam's builder, by Arnar Gauti Ingason
(MIT), kept in gleam/nix-gleam/ with its licence and a note of where
and at which commit it came from. It fetches each Hex dependency on
its own with the checksum gleam already wrote into manifest.toml, so
there is no hash in flake.nix to keep up to date. It replaces a fixed
output derivation that ran gleam deps download, which had two faults:
its hash had to be changed by hand with every dependency, and its
output was not the same bytes from one run to the next, gleam writing
packages.toml in hash-map order. A machine that had built once reused
its output and never noticed; a fresh one would have failed the hash
at random. The programs it builds give the same output as before, for
every document-language of the markup document samples.

gleam/flake.lock: flake.nix names the nixpkgs channel url, but the
lock still recorded the github pin it replaced, so every build
rewrote it. The lock as nix rewrote it is the one built and checked;
it moves nixpkgs forward to 26.11pre1088030.

README:

- the two earlier readers, sisudoc-ssp-reader and
  sisudoc-ocda-db-reader, are named as what they now are: published
  frozen, the record of readers written from the specification
  alone, their work done by ocda-read;
- the documents the checks run over are the markup document samples,
  named with their repository and the spine flags that build the
  artefacts; the claims say "every document-language" rather than a
  count, which changes as the samples do; the same in gleam/README.md,
  which also describes the nix-gleam build and names the
  specification's documents as .md, as they are;
- "one implementation per language", where it said "one reader";
- a short section on building, a link to the specification, the
  format as 2.1, and the licence.

claude-org/: the code, the READMEs and the build files are kept as
files from here on, and org is kept for setup. config-lib.org there
tangles only .gitignore and the COPYRIGHT files, and the version and
header file there has its two local git+file blocks taken out. Both
are proposed replacements for their org/ counterparts; org/d-lib.org
and org/gleam-lib.org, which tangled the code, retire with them.

.gitignore admits a LICENSE below the root, for nix-gleam's.

Checked: D and Gleam build; with the writer at the same point, over
the markup document samples: the four dumps (D and Gleam, .ssp and
.ocda.db) identical for every document-language, ocda-read the same
from every form and in both languages, and D == Gleam in all five
formats.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>dlang flake.nix touch</title>
<updated>2026-10-22T15:02:40+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-10-22T15:02:40+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=f03d6ef7ee91260e664e19c3a74dac5739f6442e'/>
<id>f03d6ef7ee91260e664e19c3a74dac5739f6442e</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>README</title>
<updated>2026-10-22T15:01:35+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-10-22T14:57:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=437d205259a15f795c19e8b9ba47348e1042f6cc'/>
<id>437d205259a15f795c19e8b9ba47348e1042f6cc</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>README: the reader in D and Gleam</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-30T00:04:55+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=4dd9b15d2210e5661f9261acf113c22d4372ffc5'/>
<id>4dd9b15d2210e5661f9261acf113c22d4372ffc5</id>
<content type='text'>
The library's reader, ocda-read: where it lives and why (src/read/,
kept out of the library in D by dub configuration, in Gleam because
nothing outside src/ is compiled), its modes in each language, and
the two properties it holds, the same output from every form and D ==
Gleam byte for byte. And the one defect it found in the library, the
database loader unable to say what it had not read. The module tables
gain ocda.inventory and the reader, the build notes the three
programs in one result.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The library's reader, ocda-read: where it lives and why (src/read/,
kept out of the library in D by dub configuration, in Gleam because
nothing outside src/ is compiled), its modes in each language, and
the two properties it holds, the same output from every form and D ==
Gleam byte for byte. And the one defect it found in the library, the
database loader unable to say what it had not read. The module tables
gain ocda.inventory and the reader, the build notes the three
programs in one result.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>Gleam: ocda-read, byte for byte the D reader's</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-30T00:04:55+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=3ce342f3644d86d68870da9d4a636b6cac6010d6'/>
<id>3ce342f3644d86d68870da9d4a636b6cac6010d6</id>
<content type='text'>
ocda-lib read, installed also as ocda-read: the modes the Gleam readers
had (the summary, --tree/--outline, --languages, --files,
--report-unknown), over the library, in src/read/, which is where
Gleam must put it. A D-only mode asked of it says so.

Its output is the D ocda-read's. For that the D reader's closing hint
no longer lists its modes, which this one has fewer of: both say that
--help lists them.

Checked over the document sample set, on the .ssp and the .ocda.db:
D == Gleam byte for byte in all five modes, 36 of 36. And the same
from the .ssp, the .ocda.db and the .zip (a .sisupod unwrapped, the
Gleam library having no zstd): the summary, --tree and
--report-unknown 36 of 36; --languages of a database == of its pod
27 of 27, and of a single-language .ssp == its database; --files
answers for a database and says why not for the others, in the same
words in both languages.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
ocda-lib read, installed also as ocda-read: the modes the Gleam readers
had (the summary, --tree/--outline, --languages, --files,
--report-unknown), over the library, in src/read/, which is where
Gleam must put it. A D-only mode asked of it says so.

Its output is the D ocda-read's. For that the D reader's closing hint
no longer lists its modes, which this one has fewer of: both say that
--help lists them.

Checked over the document sample set, on the .ssp and the .ocda.db:
D == Gleam byte for byte in all five modes, 36 of 36. And the same
from the .ssp, the .ocda.db and the .zip (a .sisupod unwrapped, the
Gleam library having no zstd): the summary, --tree and
--report-unknown 36 of 36; --languages of a database == of its pod
27 of 27, and of a single-language .ssp == its database; --files
answers for a database and says why not for the others, in the same
words in both languages.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>Gleam: inventory, what source and database hold</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-30T00:00:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=2a69213ed7c43fb4c0ab556af75c0999fe790fee'/>
<id>2a69213ed7c43fb4c0ab556af75c0999fe790fee</id>
<content type='text'>
The mirror of the D commit before it. ocda/inventory gives the
languages of any form, in the manifest's order, each with its markup
file and the sha256 of its .ssp (stored by the database, taken from
the bytes for a .ssp or a pod), and what a .ocda.db carries. The
sha256 comes from OTP's crypto through a four line Erlang FFI file,
since the standard library has none and a Hex package would be one
more fixed output hash; the whitelist admits src/*.erl for it.

The database loader counts the columns it does not read, on the
conditions the .ssp writer uses, so that it can say what it did not
understand; it could not before. The pod loader's archive, language
and text helpers become public for the inventory.

Checked: the four dumps, D and Gleam of .ssp and .ocda.db, identical
36 of 36, and with --unknown also 36 of 36; --languages of a database
== of its pod, 27 of 27.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The mirror of the D commit before it. ocda/inventory gives the
languages of any form, in the manifest's order, each with its markup
file and the sha256 of its .ssp (stored by the database, taken from
the bytes for a .ssp or a pod), and what a .ocda.db carries. The
sha256 comes from OTP's crypto through a four line Erlang FFI file,
since the standard library has none and a Hex package would be one
more fixed output hash; the whitelist admits src/*.erl for it.

The database loader counts the columns it does not read, on the
conditions the .ssp writer uses, so that it can say what it did not
understand; it could not before. The pod loader's archive, language
and text helpers become public for the inventory.

Checked: the four dumps, D and Gleam of .ssp and .ocda.db, identical
36 of 36, and with --unknown also 36 of 36; --languages of a database
== of its pod, 27 of 27.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>D: ocda-read, reader, over library, of every form</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-29T23:37:04+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=c41696e2e18d144b4e9c1e23f311dbbd67919dc2'/>
<id>c41696e2e18d144b4e9c1e23f311dbbd67919dc2</id>
<content type='text'>
One reader where there were two. sisudoc-ssp-reader read a .ssp and
sisudoc-ocda-db-reader a .ocda.db, each with its own parser; this one
is built on the library, so every mode reads a .ssp, a .ocda.db and a
.sisupod, and says the same thing of each. Its output modules are the
.ssp reader's, copied (the reader repositories are unchanged) and
moved onto the library's model, which was the reader's types renamed
and added to. From the database reader, --languages and --files.

  ocda-read [--tree|--outline|--html|--llm-text|--llm-jsonl|--images|
             --verify --image-dir=&lt;d&gt;|--languages|--files|
             --report-unknown|--validate] &lt;file&gt; [lang]

Two modes are about the file and say so: --files answers for a
.ocda.db, and --validate checks a .ssp as text and refuses anything
else. Where a file holds several languages and none is named, every
form reads the first and says so, where the database used to take one
and the pod refuse.

In src/read/ beside the library in src/ocda/; dub.json keeps it out of
the library (configurations "library" and "ocda-read"), and the flake
builds it as a package. The flake's default is now all three programs
in one bin/: built separately into the same result link, each replaced
the one before, and the pipeline's dump looked for ocda-dump where
only ocda-agree was left.

Checked over the document sample set:
- against the .ssp reader, on a .ssp, byte for byte after the first
  line (Parsed: is Read:): --tree, --html, --llm-text, --llm-jsonl,
  --images, --verify, --validate, 36 of 36; the summary differs only
  in the closing hint, which names the two new modes;
- against the database reader, on a .ocda.db: --files 27 of 27, and
  --languages 27 of 27 on language, source and digest (its id column,
  the database's own key, is left out);
- the same output from the .ssp, the .ocda.db and the .sisupod: the
  summary and --tree, --html, --llm-text, --llm-jsonl, --images and
  --report-unknown, 36 of 36; --languages, .ocda.db == .sisupod 27 of
  27, and a single-language .ssp == its database, 26 of 26.

Not reproduced, deliberately: the database reader's --outline, which
ordered the headings by their position within their own section, so
that the toc and the "(skip)" headings of other sections fell among
the body's. --outline here is --tree.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
One reader where there were two. sisudoc-ssp-reader read a .ssp and
sisudoc-ocda-db-reader a .ocda.db, each with its own parser; this one
is built on the library, so every mode reads a .ssp, a .ocda.db and a
.sisupod, and says the same thing of each. Its output modules are the
.ssp reader's, copied (the reader repositories are unchanged) and
moved onto the library's model, which was the reader's types renamed
and added to. From the database reader, --languages and --files.

  ocda-read [--tree|--outline|--html|--llm-text|--llm-jsonl|--images|
             --verify --image-dir=&lt;d&gt;|--languages|--files|
             --report-unknown|--validate] &lt;file&gt; [lang]

Two modes are about the file and say so: --files answers for a
.ocda.db, and --validate checks a .ssp as text and refuses anything
else. Where a file holds several languages and none is named, every
form reads the first and says so, where the database used to take one
and the pod refuse.

In src/read/ beside the library in src/ocda/; dub.json keeps it out of
the library (configurations "library" and "ocda-read"), and the flake
builds it as a package. The flake's default is now all three programs
in one bin/: built separately into the same result link, each replaced
the one before, and the pipeline's dump looked for ocda-dump where
only ocda-agree was left.

Checked over the document sample set:
- against the .ssp reader, on a .ssp, byte for byte after the first
  line (Parsed: is Read:): --tree, --html, --llm-text, --llm-jsonl,
  --images, --verify, --validate, 36 of 36; the summary differs only
  in the closing hint, which names the two new modes;
- against the database reader, on a .ocda.db: --files 27 of 27, and
  --languages 27 of 27 on language, source and digest (its id column,
  the database's own key, is left out);
- the same output from the .ssp, the .ocda.db and the .sisupod: the
  summary and --tree, --html, --llm-text, --llm-jsonl, --images and
  --report-unknown, 36 of 36; --languages, .ocda.db == .sisupod 27 of
  27, and a single-language .ssp == its database, 26 of 26.

Not reproduced, deliberately: the database reader's --outline, which
ordered the headings by their position within their own section, so
that the toc and the "(skip)" headings of other sections fell among
the body's. --outline here is --tree.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>D: inventory, what source and database hold</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-29T23:36:53+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=49bc3e7da926da80239b50fef39b3437adf966df'/>
<id>49bc3e7da926da80239b50fef39b3437adf966df</id>
<content type='text'>
ocda.inventory answers two questions about the file rather than the
document. languages(path), for every form alike: the languages in the
manifest's order (which every .ssp records in @source languages), each
with the markup it was made from and the sha256 of its .ssp - stored
by the database, taken from the bytes for a .ssp or a pod, and the
same. carried(path), for a .ocda.db: what it holds besides the
abstraction, by role and storage.

And the database loader now counts what it holds and the model does
not. The text loader meets every property line and counts those it
does not keep; this one asks for columns by name and so never saw the
others, and reported that it had understood everything, which is the
silence the count exists to prevent. It counts them on the conditions
under which the .ssp writes the same property, and the report is now
the same from a .ssp, a .ocda.db and a .sisupod: 36 of 36.

The pod loader's archive and language-name helpers become package
visible for the inventory.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
ocda.inventory answers two questions about the file rather than the
document. languages(path), for every form alike: the languages in the
manifest's order (which every .ssp records in @source languages), each
with the markup it was made from and the sha256 of its .ssp - stored
by the database, taken from the bytes for a .ssp or a pod, and the
same. carried(path), for a .ocda.db: what it holds besides the
abstraction, by role and storage.

And the database loader now counts what it holds and the model does
not. The text loader meets every property line and counts those it
does not keep; this one asks for columns by name and so never saw the
others, and reported that it had understood everything, which is the
silence the count exists to prevent. It counts them on the conditions
under which the .ssp writes the same property, and the report is now
the same from a .ssp, a .ocda.db and a .sisupod: 36 of 36.

The pod loader's archive and language-name helpers become package
visible for the inventory.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>model 2.1: ocn_range, the verses a poem holds</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-29T18:07:32+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=8705226727f0f26be250a07497df854e543e9645'/>
<id>8705226727f0f26be250a07497df854e543e9645</id>
<content type='text'>
The format now states, for a container, the first and last ocn of what
it holds: .ocn_range in the .ssp, ocn_range_first and ocn_range_last in
the .ocda.db. A poem is the one container, and since 2.1 it carries ocn
0 and comes before its verses.

Both fields in the model, read by both loaders, in both languages, and
written by the dump. A 2.0 database has neither column, so each loader
asks sqlite whether it is there and reads a 2.0 file as having no
ranges, which is what 2.0 meant; checked on a published 2.0 database, D
and Gleam giving the same dump.

Over the document sample set written by spine at 2.1, the four dumps
(D and Gleam, .ssp and .ocda.db) are byte identical.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The format now states, for a container, the first and last ocn of what
it holds: .ocn_range in the .ssp, ocn_range_first and ocn_range_last in
the .ocda.db. A poem is the one container, and since 2.1 it carries ocn
0 and comes before its verses.

Both fields in the model, read by both loaders, in both languages, and
written by the dump. A 2.0 database has neither column, so each loader
asks sqlite whether it is there and reads a 2.0 file as having no
ranges, which is what 2.0 meant; checked on a published 2.0 database, D
and Gleam giving the same dump.

Over the document sample set written by spine at 2.1, the four dumps
(D and Gleam, .ssp and .ocda.db) are byte identical.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
<entry>
<title>images from the source, &amp; the epub segment fields</title>
<updated>2026-10-21T23:12:17+00:00</updated>
<author>
<name>Ralph Amissah</name>
<email>ralph.amissah@gmail.com</email>
</author>
<published>2026-09-26T13:47:07+00:00</published>
<link rel='alternate' type='text/html' href='https://git.sisudoc.org/sisudoc-ocda-lib/commit/?id=d0046aa6b44e2126386351ee7c44997b6b0c46ee'/>
<id>d0046aa6b44e2126386351ee7c44997b6b0c46ee</id>
<content type='text'>
Two additions needed by the epub writer.

ocda.images: an image's bytes, from whichever input the document came
from. The model names images, with digest, byte size and pixel size, and
does not hold them (right as an abstraction is not a bundle of assets)
so something has to know where they are, and the answer depends on the
input form rather than on the output. Hence the library.

  .ocda.db   the files table, role 'image', by name
  .sisupod   the archive entry whose path ends in /image/&lt;name&gt;
  .zip       the same
  .ssp       beside it: a pod puts the abstraction in media/abstraction/
             and its images in media/image/, so the path is documented
             rather than guessed

canGiveImages says whether a source can answer at all, because a writer
has to know before it emits an &lt;img&gt;: a manifest item whose file is
absent and an &lt;img&gt; whose file is not in the manifest are both epubcheck
errors. An .ssp outside a pod has nowhere to look, and that is absence
rather than error - the document is still readable.

Reading a blob needed sqlite3_column_blob rather than _column_text: an
image read as text stops at the first zero byte, and a png's ninth byte
is zero.

And two more model fields: segment_epub and segment_epub_is. The epub
segments at heading levels 1 to 4 where the html segments at level 4
alone, so the two are different files and the abstraction records both.
Nine of the 36 documents have images; the four-way agreement still holds
at 36 of 36 each way.

(assisted by Claude-Code)
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Two additions needed by the epub writer.

ocda.images: an image's bytes, from whichever input the document came
from. The model names images, with digest, byte size and pixel size, and
does not hold them (right as an abstraction is not a bundle of assets)
so something has to know where they are, and the answer depends on the
input form rather than on the output. Hence the library.

  .ocda.db   the files table, role 'image', by name
  .sisupod   the archive entry whose path ends in /image/&lt;name&gt;
  .zip       the same
  .ssp       beside it: a pod puts the abstraction in media/abstraction/
             and its images in media/image/, so the path is documented
             rather than guessed

canGiveImages says whether a source can answer at all, because a writer
has to know before it emits an &lt;img&gt;: a manifest item whose file is
absent and an &lt;img&gt; whose file is not in the manifest are both epubcheck
errors. An .ssp outside a pod has nowhere to look, and that is absence
rather than error - the document is still readable.

Reading a blob needed sqlite3_column_blob rather than _column_text: an
image read as text stops at the first zero byte, and a png's ninth byte
is zero.

And two more model fields: segment_epub and segment_epub_is. The epub
segments at heading levels 1 to 4 where the html segments at level 4
alone, so the two are different files and the abstraction records both.
Nine of the 36 documents have images; the four-way agreement still holds
at 36 of 36 each way.

(assisted by Claude-Code)
</pre>
</div>
</content>
</entry>
</feed>
