issues/completed/7-005-select-most-recent-archive-per-type.md

Issue 7-005: Select Most Recent Archive Per Type

Priority: High
Phase: 7 (Stabilization and Polish)
Status: Completed
Created: 2026-01-30
Completed: 2026-01-30

Current Behavior

The scripts/zip-extractor.lua finds all ZIP archives in input/, categorizes them by type (fediverse, messages, notes), and extracts all of them. When multiple archives of the same type exist:

  1. All are extracted sequentially
  2. The extraction_paths dictionary overwrites previous entries: summary.extraction_paths[archive.type] = extraction_path
  3. The last archive processed (by filesystem order) is used, not the most recent by date
  4. No warning is shown about which archives were skipped

Example: If input/ contains:

  • similar-different.zip (messages, older)
  • new-messages-export.zip (messages, newer)

The extractor might use similar-different.zip simply because it was found first, ignoring the newer export.

Intended Behavior

  1. Group archives by type: Collect all archives into buckets (fediverse, messages, notes)
  2. Sort by modification time: Within each bucket, sort by file mtime (newest first)
  3. Select most recent: Only extract the newest archive from each type
  4. Warn about skipped archives: Print a yellow warning (full yellow text, not just icon) showing the full path of each older, unused archive

Example output:

🔍 Scanning for ZIP archives in: ./input
📦 Found messages archive: new-messages-export (2026-01-28)
📦 Found messages archive: similar-different (2025-12-14)
📦 Found fediverse archive: most-recent-29 (2026-01-15)

📊 Archive selection:
   messages: new-messages-export (selected)
   ⚠️  Skipped older: ./input/similar-different.zip (2025-12-14)
   fediverse: most-recent-29 (selected)

Suggested Implementation Steps

Step 1: Add modification time detection

-- {{{ local function get_file_mtime
local function get_file_mtime(file_path)
    local stat_cmd = string.format("stat -c %%Y '%s' 2>/dev/null", file_path)
    local handle = io.popen(stat_cmd)
    local result = handle:read("*a")
    handle:close()

    if result and result ~= "" then
        return tonumber(result:gsub("%s+", ""))
    end
    return 0
end
-- }}}

Step 2: Store mtime in archive info

-- In detect_archives():
table.insert(archives, {
    path = file,
    type = archive_type,
    basename = basename,
    mtime = get_file_mtime(file),  -- Add this
    mtime_date = os.date("%Y-%m-%d", get_file_mtime(file))  -- Human-readable
})

Step 3: Create selection function

-- {{{ local function select_archives_by_type
-- Groups archives by type, sorts by mtime (newest first), returns selected + skipped
local function select_archives_by_type(archives)
    local by_type = {}

    -- Group by type
    for _, archive in ipairs(archives) do
        if not by_type[archive.type] then
            by_type[archive.type] = {}
        end
        table.insert(by_type[archive.type], archive)
    end

    local selected = {}
    local skipped = {}

    -- Sort each group by mtime (descending) and pick newest
    for archive_type, group in pairs(by_type) do
        table.sort(group, function(a, b) return a.mtime > b.mtime end)

        -- First (newest) is selected
        table.insert(selected, group[1])

        -- Rest are skipped
        for i = 2, #group do
            table.insert(skipped, group[i])
        end
    end

    return selected, skipped
end
-- }}}

Step 4: Update main flow

-- After detect_archives():
local all_archives = detect_archives(DIR .. "/input")
local selected, skipped = select_archives_by_type(all_archives)

-- Warn about skipped archives (full yellow text)
if #skipped > 0 then
    for _, archive in ipairs(skipped) do
        print(COLOR_YELLOW .. "⚠️  Skipped older: " .. relative_path(archive.path) ..
              " (" .. archive.mtime_date .. ")" .. COLOR_RESET)
    end
end

-- Only extract selected archives
local summary = create_extraction_summary(selected, TEMP_DIR)

Success Criteria

  • [x] Archives grouped by type before selection
  • [x] Most recent archive (by mtime) selected per type
  • [x] Older archives in same type category show yellow warning
  • [x] Warning includes full path and date
  • [x] Only selected archives are extracted
  • [x] Extraction summary reflects actual selection

Related Files

  • scripts/zip-extractor.lua:47-61 - get_file_mtime() function
  • scripts/zip-extractor.lua:98-137 - detect_archives() function with mtime
  • scripts/zip-extractor.lua:139-170 - select_archives_by_type() function
  • scripts/zip-extractor.lua:278-303 - Main execution with selection logic
  • issues/7-progress.md - Phase 7 progress tracking

Notes

  • The yellow warning text (not just icon) helps distinguish "attention required" from "info"
  • This follows the semantic color scheme from 7-003: yellow = attention
  • Consider adding a --force-all flag to extract all archives if ever needed

Implementation Notes (2026-01-30)

Lua gsub gotcha: The string.gsub() function returns two values (string, count). When passed directly to tonumber(), the count became the "base" argument. Fixed by wrapping in parentheses: (result:gsub(...)).

Example output with multiple archives of same type:

🔍 Scanning for ZIP archives in: ./input
📦 Found messages archive: new-export (2026-01-28)
📦 Found messages archive: similar-different (2025-12-13)
📦 Found fediverse archive: most-recent-29 (2025-12-10)

📊 Archive selection:
   Total found: 3, Selected: 2, Skipped: 1
   ⚠️  Skipped older: ./input/similar-different.zip (2025-12-13)