Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/acdeveloper-sci/fortriage/llms.txt

Use this file to discover all available pages before exploring further.

The Overview is the starting point after every analysis run. It presents a high-level picture of the entire corpus — tier summary cards, two charts, and an interactive table that can be filtered and searched. Before any analysis has been run, the tab shows the upload area where you stage Fortran source files and kick off the pipeline. Once results are ready, the upload area is replaced by the full results dashboard.

Tier Summary Cards

After a successful run, five summary cards appear at the top of the page — one per tier — showing how many units landed in each severity bucket and what share of the total corpus they represent.
TierColorMeaning
CRITICALRed #D64545Highest combined complexity, coupling, and legacy debt — must be addressed first
HIGHOrange #E8874ASignificant risk; should be prioritized in the next modernization wave
MEDIUMYellow #E8C547Moderate risk; worth scheduling but not immediately blocking
LOWBlue #5B9BD5Manageable risk; straightforward to migrate when the time comes
DEAD_CODEGray #9AA5B1Units classified as unreachable — candidates for removal before any migration work
Each card displays the raw unit count alongside a caption showing the percentage of total units in that tier (e.g., 12.3% of units).

Charts

Two charts render below the tier cards and above the unit table.

Tier Distribution

A bar chart showing the unit count for each tier in canonical severity order (CRITICAL → HIGH → MEDIUM → LOW → DEAD_CODE). Use this to get an instant feel for how top-heavy the risk profile is: a corpus dominated by CRITICAL and HIGH bars requires a very different migration plan than one with most units at LOW.

CC × Fan-In Scatter Plot

A scatter plot where:
  • X-axisFan_In (number of units that call this unit)
  • Y-axisCC (Cyclomatic Complexity)
  • Bubble sizePct_Legacy (percentage of legacy-style constructs)
Units in the top-right quadrant (high CC and high Fan-In) are “core knots” — they are both intrinsically complex and heavily depended upon by the rest of the codebase. These are the most critical refactoring targets because changes to them ripple outward. Large bubbles in this quadrant compound the risk further with high legacy-construct density.

Filterable Unit Table

Below the charts, an interactive table lists every unit from report_prioritization.csv. Eleven columns are shown in display order:
ColumnDescription
PriorityTier label: CRITICAL, HIGH, MEDIUM, LOW, or DEAD_CODE
ScoreNumeric risk score computed by Forti4D (higher = more risk)
FileSource filename containing this unit
UnitFortran unit name (subroutine, function, program, etc.)
TypeUnit type: SUBROUTINE, FUNCTION, PROGRAM, MODULE, BLOCK DATA, etc.
CCCyclomatic Complexity
Fan_InNumber of other units that call this unit
Pct_LegacyPercentage of legacy-style constructs (GOTO, COMMON, EQUIVALENCE, …)
Reachability_StatusREACHABLE, UNREACHABLE, or ENTRY_POINT
Clone_StatusWhether this unit has been identified as a duplicate
StrategyRecommended migration strategy from Forti4D

Filter Controls

Three filter controls appear directly above the table:
ControlDescription
SearchCase-insensitive substring match against Unit name or File name — both columns are searched simultaneously
TypeDropdown to restrict the table to a single unit type (SUBROUTINE, FUNCTION, …) or ALL
TierDropdown to show only units in a specific tier or ALL
All three filters are combined as an AND intersection — a unit must satisfy every active filter to appear. A Showing X of Y counter above the table reflects how many rows the current filters match.
Clicking any row in the table opens the Drill-down detail panel for that unit. The unit is pre-selected in the Drill-down tab’s selector so you can navigate directly to its full metric profile.

Load Sample Fixtures

The Load sample fixtures button, shown in the upload area before any analysis is run, loads a built-in corpus of 26 Fortran units from 8 source files. This is the same fixture set used by Forti4D’s own test suite and is committed to the ForTriage repository at tests/fixtures/fixtures.zip. Loading sample fixtures is identical to uploading a real zip archive — the same extract_zip_entries path processes it, no separate loading mechanism is involved. Once loaded, a caption reads:
Sample fixtures loaded — click "Clear all" to remove them.
Click Clear all (top-right of the upload area) to remove the fixtures and return to the blank upload state.

Build docs developers (and LLMs) love