Project Import — PeopleCode Diff vs Database

Tailored Operational Context
  • Target Database:
  • Context Type:
  • Alert Severity:
  • Triggered Time:
  • Firing Context:

Project Import — PeopleCode Diff vs Database

Report ID: project-import-diff Category: Objects

Purpose

You uploaded a PeopleSoft project XML on the Project Import page, and you want to know — line-by-line — how every PeopleCode program inside that file compares to the same program in a live database.

The Project Import results page itself shows a quick Same / Different / Not in DB badge per object (powered by a content-hash compare against PSPCMTXT). That’s enough for triage. When you need to see what actually changed, run this report; it emits a full markdown document with file source, database source, and a unified diff for every PeopleCode object that differs.

The output is designed to be reviewed by hand or fed directly to an LLM — the per-object structure (file block, DB block, diff block) is what most assistants need to reason about the change.

What It Compares

This report only looks at PeopleCode-bearing object types:

  • Record PeopleCode (type 8)
  • Menu PeopleCode (type 9)
  • Message PeopleCode (type 39)
  • App Engine PeopleCode (type 43)
  • Page PeopleCode (type 44)
  • Component PeopleCode (type 46)
  • Component Record PeopleCode (type 47)
  • Component Record Field PeopleCode (type 48)
  • Application Package PeopleCode (type 58)

For each PeopleCode item in the project, the report:

  1. Reads the source from the uploaded XML (peoplecode_text node parsed at upload time).
  2. Fetches the current source from the target database’s PSPCMTXT table, concatenating multi-row PCTEXT payloads in PROGSEQ order.
  3. Normalizes both sides (line-ending normalization, trailing-whitespace strip, blank-line collapse) so cosmetic differences don’t show up as content changes.
  4. Classifies the item as Same, Different, Not in DB, No XML Source, or Error.
  5. For Different items, renders both sources and a line-level diff inline.

Non-PeopleCode object types in the same project are not included in this report — use the Project Import results page for those (it uses LASTUPDDTTM comparison, which is reliable for non-PeopleCode types).

Parameters

ParameterRequiredDefaultDescription
project_idYesThe file-store ID of the uploaded project. The Project Import results page provides a “PeopleCode Diff Report” button that pre-fills this.
max_objectsNo100Cap on how many Different objects are rendered with full source + diff. Excess Different objects are mentioned in a footer line; raise the cap if you need them all inline. Prevents the report from blowing up on projects with hundreds of changed PeopleCode programs.
include_sameNofalseSet to true to append a table listing every PeopleCode object that matched. Off by default since this is usually the noise, not the signal.

The target database is the standard psLens database selector on the Reports page — the report runs against whichever database you have selected.

Output Structure

The report is a single markdown document with these sections in order:

  1. Header — project name, source DB (from the XML export), target DB, export date, exporter ID, uploaded filename, upload time.
  2. Summary — counts of Same / Different / Not in DB / No XML Source / Errors.
  3. Different — one section per Different object containing:
    • File source — code fence with peoplecode language hint.
    • Database source — code fence with peoplecode language hint.
    • Line diff (file → DB) — code fence with diff language hint; lines unchanged on both sides appear with two-space prefix, removed-from-file lines with -, added-in-DB lines with +.
  4. Not in Database — one section per item, with the file source rendered. These are typically programs the project would create on import.
  5. No XML Source — table of programs the DB has but the XML didn’t inline. Usually means the project listed the PJM entry without the PCM payload.
  6. Errors — table of programs whose DB query failed, with the underlying error.
  7. Same (optional, include_same=true) — table of matching programs.

Tips

  • Feed it to an LLM. The structure (file block + DB block + diff block per object) is what assistants need to reason about a change. Download the markdown and paste into your assistant of choice, or use the Download .md button on the report run page.
  • Pair with the Cross-Database Comparison report. project-compare covers two databases; this one covers file vs database. Different problems.
  • Run before a project import. Knowing exactly what will change in PeopleCode terms is the question this report answers — not “what does the timestamp say” but “what code lines will end up different.”