Fixes and docs (#73)

* Fix inserted rows count to avoid returning 0 when the count is not returned by the backend
* Add zooming capability in `xml2db serve`
* Write documentation for CLI usage
* Several documentation editing

---------
Co-authored-by: Claude <noreply@anthropic.com>

Fix DuckDB bulk_insert failing when quoted CSV cells fall outside sni…

…ffer sample (#64)

* Fix DuckDB bulk_insert failing when quoted CSV cells fall outside sniffer sample (#62)

DuckDB's read_csv sniffer only examines the first ~20 480 rows; if none are
quoted, it sets quote=(empty) and then errors on any later quoted cell with a
column-count mismatch.  Passing quote='"' explicitly bypasses auto-detection.
Adds a regression test that inserts 25 000 plain rows followed by one row
whose value contains a comma (triggering csv.writer quoting).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Bump version to 0.13.3

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Add explicit escape='\"' to read_csv alongside quote='\"'

Makes the RFC 4180 doubling behaviour explicit rather than relying on
DuckDB defaulting escape to the same char as quote.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

Fix KeyError when xs:choice branches share an element with same name …

…and type (#60)

* Fix KeyError when xs:choice branches share an element with same name and type

In add_relation_1/add_relation_n, skip duplicate relations where a field with
the same name and same XSD type already exists. Previously, self.fields (a list)
accumulated both occurrences while self.relations_1/n (dicts) only kept the last,
causing a KeyError in simplify_table when the first occurrence was elevated and
the second tried to look it up.

Add deliveryType to the orders crash-test schema to cover this case, with
order3.xml exercising both the sequence branch (from+to) and standalone branch (to).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Bump version from 0.13.1 to 0.13.2

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

Document and test multiprocessing pattern for concurrent XML loading (#…

…58)

Adds a test_multiprocessing.py test (parallel parse + serialised DuckDB writes via multiprocessing.Lock, with XML roundtrip content assertion) and a matching example in the API overview docs. Each worker creates its own DataModel with a unique temp_prefix so temp tables never collide. Bumps version to 0.13.1.

Add DatabaseDialect and implement identifiers truncation (#54)

* Added a DatabaseDialect class hierarchy (base, postgresql, mssql, mysql, duckdb), one subclass per backend, to abstract db specific logic and avoid conditionnals scattered in the code
* Implement truncation of database identifiers, because long names was an issue frequently found especially with postgresql. Each db dialect sets its own limit
* Remove Python 3.9 support (EOL) and add Python 3.14 to the test matrix
* Bump version
---------

Co-authored-by: cre-os <opensource@cre.fr>

Allow xsd:any types (#52)

* Allow xsd:any types by dropping corresponding data

* Bump version

---------

Co-authored-by: cre-os <opensource@cre.fr>

Fix models tree repr (#33)

Add a test for stable source and target tree repr

Co-authored-by: cre-os <opensource@cre.fr>

Move hashing after conversion (#22)

Changed the way hashes are computed in order to compute them after converting data types, so that data which is at the identical in the database result in identical hashes.

---------

Co-authored-by: cre-os <opensource@cre.fr>

Fix integer enumeration parsing (#14)

Schema parsing failed with integer enumeration, because we tried to get field size from the enumeration elements `len`, which does not work for `int`

Switch to lxml XSD validator (#12)

* Add imported XSD schema to test cases
* Switch XML validator to lxml
* Improve docs