OvitalMap Parcel CSV

Use the bundled Python scripts through JSON stdin/stdout. Do not recreate coordinate conversion, code allocation, CSV generation, or archive updates manually.

Input

Set OVITALMAP_WORKSPACE to the directory where exports and archives should be stored. If it is unset, the scripts use the current working directory.

Each parcel is a JSON object:

{"vertices":[[114.13472,22.50422],[114.13564,22.50411],[114.135,22.503]],"provider_name":"Survey Team","official_id":null,"altitude":[]}

Use WGS84 longitude, latitude order. For a batch, keep one country or region per pipeline run. The pipeline assigns stable parcel_ref values (P01, P02, and so on).

Workflow

1. Extract and show the source coordinate text. When the source is an image, preserve the transcription for review.
2. Convert decimal, DMS, or UTM coordinates with scripts/coordinate_converter.py. Supply the coordinate format; for decimal input also supply order, and for UTM supply zone and hemisphere.
3. Show the resulting WGS84 vertices and obtain explicit confirmation of the coordinates and provider names. Do not write exports or archives before confirmation.
4. Obtain the ISO 3166-1 alpha-2 country or region code from explicit context. Ask when it is uncertain.
5. Run python3 scripts/parcel_pipeline.py --step 1 with parcels, country_code, and optional date in YYMMDD form. Retain the returned run_id. On needs_input, request only the fields listed in required_input.
6. Continue with the same run_id: run Step 2b to classify archive hits and new parcels, then Step 2 to assign codes to new parcels.
7. Show the proposed codes. After the user approves them, pass confirmed_codes: true to Step 3.
8. Deliver every path in result.exports in its returned order. Boundary files import into OvitalMap as tracks (轨迹); vertex files import as labels (标签).

Use --step all only when the user has already confirmed the complete input and explicitly approved automatic code acceptance with "confirmed": true and "auto_accept_codes": true.

Parcel codes

Prefer a confirmed official registration, cadastral, or permit identifier:

{CC}-{OFFICIAL_ID}

When no official identifier is available, assign a stable archive code:

{CC}-{YYMMDD}-{SEQ}

The archive determines the next three-digit sequence. Do not use unknown, invent a descriptive parcel name, or silently replace a generated code later.

Export modes

• boundary is the default and produces one track CSV per parcel.
• vertices produces one label CSV per parcel only when explicitly requested.
• both produces both formats only when explicitly requested.

Preserve the chosen mode throughout a multi-step run.

Safety and consistency

• Treat error, blocked, and needs_input as stopping conditions.
• Never guess the country, provider, altitude, official identifier, UTM zone, or hemisphere.
• Reuse an archive hit's parcel code and metadata; do not append it again.
• Resolve batch conflicts by parcel_ref; identical coordinates do not prove two submissions are the same parcel.
• Do not replace a non-empty cadastral identifier without explicit confirmation.
• Preserve vertex order; the scripts close polygons when needed.
• Export each parcel separately and preserve submitted order.
• The scripts own filenames, CSV headers, encoding, archive schemas, and defaults.

Archive operations

Use scripts/archive_manager.py with one of these actions: scan, check_duplicate, extract_single, update_cadastre, backup, or correct.

References

• Read references/csv-contract.md when checking or changing OvitalMap CSV compatibility.
• Read references/workflow-contract.md when handling pipeline states, confirmations, or errors.
• Read references/interaction-and-edge-cases.md for duplicate submissions, archive hits, corrections, or multi-parcel batches.