À propos des fichiers XML

XML (XML Document · Format de document)

Lancement 1998 Par W3C (Tim Bray, Jean Paoli, C. M. Sperberg-McQueen)

XML Document (.xml) est un format de fichier pris en charge par KaijuConverter pour l’archivage, l’échange professionnel et la conversion multiplateforme.

Famille

Documents & Text

Extension

.xml

Type MIME

application/xml, text/xml

Utilisable comme

Entrée

L'histoire

COMMENT XML
A VU LE JOUR.

1986

1998

XML — Extensible Markup Language — was ratified by the W3C in February 1998 as a simplified subset of SGML (which had been an ISO standard since 1986). The design goal was elegant: all the power of SGML’s customizable tags, but in a format simple enough that anyone could write a parser over a weekend. For the five years after 1998, XML was genuinely exciting — it underpinned SOAP, RSS, XHTML, DocBook, SVG, and countless vertical industry standards (HL7 for healthcare, FIXML for finance).

Then JSON happened. For most lightweight API work JSON was 30% smaller, vastly easier to parse, and worked natively in browsers. XML retreated into strongholds where its strengths mattered: schema validation (XSD), digital signatures (XML DSig), document semantics (DocBook, TEI), vector graphics (SVG), and enterprise integration (SOAP, XBRL). Those niches are still alive and standardized — XML is not dead, it just handed the casual lane to JSON.

CURIOSITÉS ET
ANECDOTES.

XML was designed by an 11-person working group in roughly 20 weeks — unusually fast for a W3C standard.

An XML document that follows only the syntax rules is "well-formed"; one that also matches a schema is "valid".

SVG, DocBook, TEI, MusicXML, GPX, and countless other vertical formats are all XML under the hood.

XML namespaces were added in 1999 to let documents mix tags from multiple schemas without name collisions.

The fastest pure-Java XML parser (Aalto) can parse 1 GB of XML in under 10 seconds on commodity hardware.

POINTS FORTS ET
LIMITES.

Points forts

Self-describing tags make documents semantically rich and human-readable.
Schema validation (XSD, RelaxNG, DTD) catches structural errors before they hit production.
Namespaces let unrelated vocabularies coexist in one document.
Mature ecosystem: XPath, XSLT, XQuery, DSig, XML Encryption all layer on top.
Preferred format for regulated industries (healthcare, finance, government) that require validation and audit trails.

Limites

Verbose — file sizes are typically 2-5× larger than equivalent JSON.
Parsing is expensive compared to JSON, especially for small messages.
Namespaces and DTD processing have historically been security attack vectors (XXE, billion-laughs).
Learning curve is steep for the advanced stack (XSLT, XSD, XPath).
Most developers today prefer JSON; XML tooling is aging.

Tailles et poids typiques

Small config file

1-10 KB

RSS feed

10-200 KB

Enterprise SOAP message

50 KB - 2 MB

Wikipedia XML dump

~20 GB compressed, ~100 GB raw

Spécifications techniques

MIME types: application/xml, text/xml
Extensions: .xml, plus format-specific (.svg, .xsd, .xsl, .rss, .atom)
Standard: W3C XML 1.0 (Fifth Edition, 2008)
Character encoding: UTF-8 or UTF-16 (declared in prolog)
Related: XSLT, XPath, XQuery, XSD, XML DSig

CONVERTIR DEPUIS
XML

xml pdf xml docx xml odt xml txt xml md xml html xml epub xml rst

Cas d'usage courants

Data interchange, configuration files, API responses, structured data.