com.yaktool/labels

Lay out FNSKU labels as an exact-physical-size PDF for 30-up US Letter (Avery 5160 geometry), A4 24-up, or 2"×1" / 2.25"×1.25" thermal rolls, and write it to `out`. Code 128 bars are aligned to the target printer dot grid (203/300 dpi) — X-dimension is always a whole number of printer dots, which is what makes the printed label scan reliably. EVERY ROW IS VALIDATED FIRST: not 10 characters, lowercase, punctuation → the row is rejected and reported (LBL-F,* findings), never silently printed; a 12–13 digit input is called out as a UPC/EAN pasted into the wrong column. X00 and B0 (FNSKU=ASIN) prefixes are both accepted — the "always X00" rule of thumb is community consensus, not official grammar, and the tool says so. TITLES ARE PRINTED WITH A STANDARD PDF FONT (WinAnsi): characters outside it (CJK, emoji) are replaced or dropped WITH DISCLOSURE (LBL-T01) — relay that disclosure, never present the label as containing text it does not. THE PDF MUST BE PRINTED AT 100% / "Actual size" — fit-to-page scaling breaks bar widths and the label grid (LBL-P01 is attached to every result; tell the user). Pass `calibrationOnly: true` to get a ruler test page (1" must measure 25.4 mm) instead of labels. startCell (0-based) starts printing at a later position on a partly used 30-up/24-up sheet. Returns { summary{requested,printable,rejected,labels,pages}, items[{input,normalized?,valid,findings[]}], findings[], wrote }.

Lay out GS1-compliant logistic labels (A6/4"×6" SSCC pallet label, or A5/6"×8" with trade item data) and write them to `out` — as an exact-physical-size PDF, or as ZPL when `out` ends in .zpl (bar-by-bar geometry identical to the PDF; send the file straight to a Zebra-compatible thermal printer). THE LAYOUT AND NUMBERS ARE THE OFFICIAL ONES: three building blocks with the SSCC barcode always lowest, 31.75 mm bars (that is the official bar height — the "32 mm" people remember is a carton placement distance), X-dimension pinned to whole printer dots inside the 0.495–0.940 mm window (4 dots at 203 dpi, 6 at 300 dpi), ≥10X quiet zones, HRI below each symbol with parentheses around AIs. Every rule cites the GS1 General Specifications or the GS1 Logistic Label Guideline (both free official PDFs). VALIDATION IS A GATE, NOT A SUGGESTION: SSCC check digit verified (17 digits auto-complete it, and the result reports the completed SSCC); GTIN (02) and count (37) are only valid together; compact formats print the SSCC only (LGL-010 warning when trade fields are ignored); one symbol carries at most 48 data characters; dates are YYMMDD where day "00" legitimately means end-of-month. Rejected rows are reported with LGL-* findings, never silently printed. X-DIMENSION IS NEVER SHRUNK below the compliant window to squeeze data in — over-long data is an error (LGL-011/012) with a fix suggestion. Returns { summary{requested,printable,rejected,labels}, items[{input,sscc?,valid,findings[]}], findings[], wrote }.

Parse ZPL II text, render every ^XA…^XZ label block, and write an exact-physical-size PDF to `out` — proof a label file before it costs a strip of label stock. SUPPORTED SUBSET: label framing (^XA ^XZ ^PW ^LL), positioning (^FO), boxes (^GB), text (^A0/^CF/^FD) and Code 128 barcodes (^BC modes N and A with >; / >8 escapes) — bars are drawn by the same encoder as this suite's generators, cross-checked against two independent barcode libraries. HONEST COVERAGE, NEVER GUESSES: commands outside the subset (other symbologies like ^B3/^BQ/^BX, images ^GF, rotated fields, firmware check-digit modes U/D) are SKIPPED AND LISTED in `coverage.ignored` and ZPV-* findings — a wrongly-guessed barcode is worse than a missing one. Relay the coverage report; never present a partial render as complete. Labels without ^PW/^LL get an inferred canvas (disclosed via ZPV-004). Returns { labels, coverage{rendered,ignored}, findings[], wrote }.