Fixing Russian Legal Code Article Number Parsing

[mikhashev]

The Problem

When importing Russian legal documents from official government websites, we encountered a puzzle: the article numbers were ambiguous. The number 531 could mean article 531, or it could be sub-article 53.1. The number 12316 could be article 12316, or sub-article 123.16. Even more confusing, 1061 appeared right after article 104, with no 105 or 106 in sight.

The Mystery

Russian legal codes have a complex structure. Each code has main articles (Civil Code Part 1 has 1-453), but also sub-articles like 53.1, appendices like 12316-1, and sometimes entire articles get deleted by law while their sub-articles remain.

The government websites don't use dots or special formatting. Everything is run together. Our import system was getting confused, leaving some articles as 531 when they should be 53.1, or breaking 12316-1 into 123.1.6-1 instead of 123.16-1.

The Solution

We built a three-part detective system:

1. Context Clues - Look at neighboring articles. If 531 appears between 53 and 54, it must be 53.1.

2. Known Boundaries - Each code has valid ranges. Civil Code only goes to 453, so 601 must be 60.1.

3. Deleted Article Handling - When articles 105-106 were deleted by law, their sub-articles remained. The system learned that 1061 after 104 means 106.1 - a sub-article of a deleted main article.

The Result

All 23 Russian legal codes (~6,700 articles) now import correctly:

• 531 becomes 53.1
• 12316 becomes 123.16
• 12316-1 becomes 123.16-1
• 1061 becomes 106.1
• 602 becomes 60.2

The system can now handle the complex and ambiguous encoding used in official Russian government sources.