Conversation
Mearman
force-pushed
the
feat/100-percent-mutation-ooxml.js
branch
5 times, most recently
from
September 16, 2026 01:00
3c46d9b to
856c7dc
Compare
…code buffer sizing Adds direct coverage for bytesToBase64/base64ToBytes across every input-length remainder (0, 1, 2 bytes past a full 3-byte group), the invalid-base64 throw for each of the two positions a malformed character can occupy in a 4-character group, and whitespace stripping before decode. base64ToBytes now builds its output as a plain number[] converted via Uint8Array.from rather than pre-sizing a Uint8Array from a `len * 3 / 4` estimate: that estimate is only ever an upper bound, so any formula that never under-counts is behaviourally identical to any other once the result is trimmed to its real length -- removing the sizing arithmetic as an AST node rather than leaving an unobservable estimate for a mutation to hide behind.
…lder scaffolding
Adds direct coverage for buildXml across every XmlNode variant (text,
comment, cdata, pi, declaration, attribute-less and attributed
elements, nested children, multiple root nodes) and for
assertBuiltString's own throw, extracted from buildXml so the "did the
builder return a string" guard is directly testable with a non-string
literal rather than left uncovered forever (XMLBuilder, given this
module's fixed options, never actually returns anything else).
Simplifies two spots verified directly against fast-xml-parser to be
unobservable: a processing instruction's and a declaration's own child
array is never rendered by the builder under this configuration (`{
"?custom": [{ "#text": "x" }] }` and `{ "?custom": [] }` build to the
byte-identical `<?custom?>`), so neither carries a value the builder
ever reads; and an element's own `:@` attributes object is set
unconditionally rather than gated on whether any attribute exists,
since an empty `:@": {}` builds identically to the key being absent
and parseAttributes already reads both back to the same empty array.
Exports and directly unit-tests every one of parseXml's own structural guards and error paths (isRecord, isUnknownArray, asString, parseNodes, parseNode, parseAttributes, scalarText) against synthetic fast-xml-parser-shaped input: a node that is not an object, a node with no tag key or more than one, an attribute value or scalar-text wrapper of the wrong shape. Real fast-xml-parser output never produces these malformed shapes, so none of these branches was ever exercised through parseXml's own public entry point alone.
…variant Adds direct coverage for isXmlNode's own structural guard across non-record inputs (null, an array, a primitive -- each a distinct branch of typeof/null/Array.isArray that real Zod-validated input never separately exercises), every XmlNode variant's own required fields, malformed attribute entries, and a recursive check that a child element's own children are validated the same way rather than only its own direct fields.
…edundant bounds check Adds direct coverage, via packageFromEntries's own xml/binary classification, for a UTF-8 BOM prefix (alone and combined with leading whitespace), every individual whitespace byte the format permits, a run of several in a row, an all-whitespace part with no non-whitespace byte at all, and a part whose first three bytes only partially match the BOM (isolating each of the three signature bytes' own necessity) -- none of which any existing test exercised. Drops looksLikeXml's own `bytes.length >= 3` BOM guard: it is provably redundant given how out-of-range Uint8Array indexing behaves -- an index at or past a real array's own length always reads `undefined`, which can never equal a real BOM byte, so a short array already fails the byte-by-byte comparison on its own. The main scan loop is likewise rebounded on `bytes[i] !== undefined` rather than a separately tracked `i < bytes.length`, for the identical reason.
…ant bounds check Adds direct coverage for sniffImageFormat across every recognised signature (PNG, JPEG, both GIF header versions), near-miss prefixes that diverge partway through or on the final byte, and SVG detection by its own XML-prolog and bare-root-tag spellings, leading whitespace before either, and the 1024-byte sniff window's own boundary (a real '<svg' tag placed well past the window must not be found there). Drops startsWith's own `bytes.length < signature.length` guard: it is provably redundant given how out-of-range Uint8Array indexing behaves -- an index at or past a real array's own length always reads `undefined`, which can never equal a real signature byte, so a shorter array already fails the byte-by-byte comparison on its own.
Adds direct unit coverage for relsPathFor (a slash-free part path, and a nested one where only the LAST slash may split it) and resolveRelTarget (a package-rooted target, a relative target against both an empty and a real subject directory, a '../' segment popping the enclosing directory, a '.' segment, and a doubled-slash empty segment) -- neither function was reachable from any existing test except through a much larger relationship-resolution fixture that never varied these specific shapes.
… redundant date checks Adds direct coverage for serialToIsoTime/serialToIsoDateTime's own non-finite and negative-serial rejections, and for utcMsOfCalendarDate's own year/month rollover rejections -- including a day value large enough to roll a whole leap year forward, the one shape that makes the year check's own necessity observable (the public isoDateToSerial entry point never passes a day outside 0-99, which alone never triggers it). isoDateOfDayCount now switches on the sign of the offset from the phantom leap day rather than pairing an equality check (excluding day 60 itself) with a separate `<` comparison against the identical threshold: with 60 excluded by the `0` case, the remaining two cases are Math.sign's only other outputs, leaving no inequality boundary for a mutation to hide behind. utcMsOfCalendarDate drops its own third, day-level equality check: Date.UTC(year, month-1, day) maps onto exactly one real calendar date, so whenever a re-read year and month both already match what was asked for, day is necessarily inside that month's own valid range and is therefore already forced to match too (confirmed by exhaustive search over every realistic year/month/day combination) -- a third check here could only ever restate a fact the first two already guarantee.
…space split Adds direct coverage for parseSqref (absent/empty input, a single bare cell, a real span, several ranges, a malformed token skipped among well-formed ones), formatSqrefRange (bare cell vs. row-only vs. column-only vs. full spans), and formatSqref's own join -- none of which this shared helper had a dedicated test file for at all. Simplifies the token split from `/\s+/` to `/\s/`: splitting on each individual whitespace character rather than a run of them only ever inserts extra empty strings between adjacent whitespace characters, which the loop's own `token === ""` skip already discards, so both forms produce the identical final token list regardless of how many consecutive whitespace characters separate two ranges.
… directly Adds a dedicated test file for the xlsx rule-residue helpers: capturing zero, some, and every attribute as unmanaged, and reading residue back for an absent source, a wrong-format source, a source that fails to parse as exactly one element, and one whose tag mismatches the expected rule kind -- none of which had direct coverage before.
Adds a dedicated test file: an absent sharedStrings part reads back as exactly an empty array (not a placeholder value), multi-run <si> entries concatenate in document order, and SharedStringTable assigns sequential indices while deduplicating a value interned twice.
…re no tables A plain property read cannot distinguish a genuinely absent key from one spread on with an explicit undefined value -- both read back as undefined. Adds an Object.hasOwn check alongside the existing toBeUndefined() assertion so readXlsx's own conditional spread is actually exercised, not just its value.
… directly Adds a dedicated test file: a sheet with no table relationship at all reads no definitions, a non-table relationship among several is skipped in favour of the genuine table one, and a table part missing its own name or ref attribute is skipped rather than promoted with a missing field.
…at all Adds a case where neither of an image's own neighbours is a paragraph (two more images either side), which no existing fixture in this file exercised -- every prior case had at least one paragraph candidate, matching or not.
…r patterns' own absent key Adds a "none" w:fill and a "none" w:color case (only "auto" was previously exercised for either), and strengthens the existing single-colour pattern tests with an Object.hasOwn check: a plain toEqual cannot distinguish an omitted foregroundColor/backgroundColor key from one spread on with an explicit undefined value, so a genuinely one-sided pattern read needs the stricter check to prove the other key is truly absent.
…nt, and reply linkage Adds a dedicated test file: threadedCommentId's own uppercase-hex formatting (a counter of 10 exercises the digit-vs-letter distinction 0-9 alone cannot), sequential ids increasing across two separately commented cells (not just within one thread), a reply immediately following its own root with the root's real id as parentId while the root itself carries none, and the threaded-comments root's own declared namespace. Exports threadedCommentId, previously module-private, purely for this direct coverage.
… a dead type-narrowing check Adds a test at exactly the half-point tolerance boundary (not just comfortably inside it), and four tests each isolating one dimension's own necessity in pageSizeToPaperSizeCode's Letter/A4 checks (a width match with a mismatched height, and vice versa, for both page sizes) -- none of which any existing test distinguished from the other. parseUniversalMeasureToPt no longer runs an `amountRaw === undefined || unit === undefined` check after a successful regex match: neither of UNIVERSAL_MEASURE_RE's two capture groups is optional (neither has a trailing `?`), so a successful match always populates both -- TypeScript's own RegExpExecArray typing just cannot express that a specific pattern's own groups are mandatory. Non-null assertions state that directly instead of a runtime check no real regex match can ever fail.
…and numeric level ordering
Adds a w:startOverride whose own ilvl names a level the base
abstractNum never defined (must be skipped, not fabricated), a
declared-namespace assertion for the built w:numbering root, and a
level ordering case proving ilvl sorts numerically ('10' after '2'),
none of which the existing round-trip-only fixtures distinguished from
a passing but coincidentally-correct result.
readEmbeddedOoxmlPayload's outer catch swallows a wrongly-detected flavour's own read failure exactly as gracefully as a genuinely undetected one, so testing hasDocxBody/detectFlavour only through that public entry point cannot tell "correctly found no flavour" apart from "wrongly matched one, then threw reading it" -- both produce the same undefined result. Exports both functions and adds direct coverage: a w:body present/absent, and each of the three entry-part flavours detected (or none) independent of the read that would follow.
…cell ContentSheetCellSchema requires displayText, absent from the plain number-cell literals comments-write.test.ts built by hand -- caught by tsconfig.node.json's own typecheck (which includes test files, unlike the base tsconfig.json a plain tsc run checks). Introduces a numberCell helper that always sets it alongside the numeric value.
…tes.length Uint8Array.prototype.subarray already clamps its end argument to the array's own length, so requesting SVG_SNIFF_WINDOW bytes from a shorter buffer already yields exactly the bytes that exist -- the Math.min was never observably different from omitting it.
…hape A value shaped exactly like a valid element (tag/attributes/children all present) under an unrecognised type name must still fall through to the final `return false` -- nothing previously drove the value into the "element" arm by an unrelated type name alone.
…wards A comment thread with one reply, followed by a second cell's own comment, needs the second root's id to continue at 2 -- a reply-loop increment that ran backwards would instead collide it with the first cell's own root id.
A distractor relationship whose type is not the table relationship type, but whose target happens to be a genuinely well-formed table element (name and ref both present), must still be skipped -- the existing distractor test's target failed the name/ref check anyway, so it could not by itself distinguish the type guard from an absent one.
… guard When indexOf finds no 'T', the date half slices to length iso.length - 1 and the time half to the whole iso.length characters. ISO_DATE_PATTERN and ISO_TIME_PATTERN are anchored to exactly 10 and 8 characters respectively, so matching both at once would need iso.length to be both 11 and 8 -- impossible. With no separator, at least one half always fails to parse, so the existing undefined fallthrough already covers it with no separate check needed.
insertConstructMarkers's own "extents.length === 0" early return produces the same array content the main loop already builds for an empty extent list, and isBlockScopedHalf's trailing calc no longer needs its "lastContentIndex === -1" shortcut: position is guaranteed non-negative by the guard above it, so "position > lastContentIndex" already evaluates true on its own whenever lastContentIndex is -1. Both readCheckboxState and readOnOff drop the identical "val === undefined ||" shortcut for the same reason -- undefined already satisfies every one of the three !== checks that follow it.
…pairing gaps Adds direct unit coverage for indexParagraphContent's content-bearing classification, isBlockScopedHalf's leading/trailing edge cases via synthetic ParagraphContentIndex objects, runRangeMarkerExtents' malformed start/end pairings and out-of-order run positions, compareExtents' startIndex-over-order sort priority for crossing extents, and every w:/w14: spelling fallback across readContentControlDescriptor and readFormControlDescriptor's checkbox, dropdown, and gallery reading.
…tests The two boundary tests introduced in the prior commit compared a half's runPosition rather than its actual array position (index.elements.indexOf), so both silently exercised the wrong slots and left the </= boundary mutants on isBlockScopedHalf's leading/trailing calc alive. Pads the elements array with filler so each half lands at the exact index the test means to probe, and pins the other half's own block-scope verdict via an unrelated, mutant-stable condition so the pair's AND genuinely hinges on the one boundary under test.
…checks Number.isInteger(min)/(max)/(r) is always true or NaN given each value's own Number.parseInt provenance, and min >= 1 (or r >= 1) already rejects NaN unaided, so the isInteger guards were checking exactly what the numeric bounds already reject. A "max >= min" guard on a declared column range is equally unnecessary: columnWidthPt's own lookup only ever matches a range via "index >= min && index <= max", which an inverted range can never satisfy for any index, so admitting one unguarded is exactly as inert as rejecting it. Introduces parseIntAttr to read min/max/r directly as NaN-when-absent, replacing the "attr(..) ?? \"\"" placeholder Number.parseInt needed only to satisfy its own string parameter -- every string that could stand in for "absent" parses to NaN just the same, so the placeholder's own text was never an observable choice.
…ck ternaries
Number(undefined) is already NaN, and every one of these ternaries fed
that NaN straight into an isFinite check that already degrades it to the
same fallback (0, or DEFAULT_ROW_HEIGHT_PT) an explicit NaN branch would
produce -- readAnchorChild's own "empty string" arm is the same story,
since Number("") is 0, itself already finite and thus already the
function's own fallback value. parseIntAttr's identical-shaped ternary
stays: Number.parseInt requires a genuine string argument, so the
"undefined" branch there is load-bearing for the type system even though
it is provably behaviourally equivalent to the value parsing would
already produce.
…As sizing Adds synthetic-package tests for xlsx drawing-anchor geometry: a malformed column range (min below 1) falling back to the default width, a covering range's own declared width winning over a wider range with no width at all, a real sheetFormatPr defaultRowHeight overriding the built-in default, a declared row's own height taking precedence over that default, a malformed row (r below 1, or an unparseable ht) falling back to the default height, and editAs defaulting to twoCell (to-marker sizing) versus reading an explicit oneCell (own transform-extent sizing). Also names the payload sheet from the graphic frame's own xdr:cNvPr/@name in the existing chart graphic frame test, rather than leaving it implicit.
…gaps Adds synthetic-package tests for the chart-graphic-frame reading path that the picture-anchor fixtures never exercised: a graphicData whose uri names something other than a chart, a graphic frame with no xdr:cNvPr at all, one whose cNvPr carries no name attribute, a worksheet whose rels list an unrelated relationship type before the real drawing one, and a picture-only drawing asserting embeddedObjects stays absent. Also adds marker-field tests distinguishing a genuinely nonzero rowOff from colOff and a numeric marker value from a non-text sibling node, a column-range test proving a range never applies below its own declared min, and an absoluteAnchor position landing exactly on a column boundary. Drops the now-provably-redundant "r >= 1" guard on declared row heights: rowHeightPt is a direct Map.get on the caller's own index, never a range test, so a malformed row lands at a key no legitimate query can ever reach, unlike the analogous column-range check this guard was modelled on. Reads editAs directly against "oneCell" rather than through an intermediate default, since twoCell and an absent attribute are already indistinguishable to that comparison. Documents emptyWorksheet's own tag as unobservable to its sole caller. Rewrites chartCells to read a table cell's single run directly instead of joining a general multi-block, multi-run shape neither this file's only producer (labelCell) nor any real chart cache ever populates with more than one of either.
…ixed package-scaffolding XML buildDrawing's zero offsets, rect preset, and distT/B/L/R attributes, and the fixed _rels/.rels relationships, [Content_Types].xml Default/Override entries, and styles.xml docDefaults/Normal scaffolding were never asserted against their literal values: readDocxContent doesn't read most of them back, so a round-trip assertion alone can't catch a mutated literal. These tests parse the written XML directly and check every fixed attribute value.
…tion gaps Adds direct synthetic-package coverage for sheetFormatDefaultRowHeightPt, readColumns, and readRows: default row height fallback, 1-based min/r lower bounds, the 1-based-to-0-based index subtraction, and hidden flags. Covers deriveDisplayText/resolveNumericValue's exact per-kind displayText output (dateTime, percentage, false boolean, symbol-only currency), readCellValue's boolean parsing and NaN handling, and merged-range colSpan/rowSpan arithmetic anchored away from row/column 0 so subtraction and addition mutants actually diverge. Adds a hasOwn() helper and uses it wherever a test needs to prove a key is genuinely absent from a ContentSheetCell/ContentSheet, since toBeUndefined() cannot distinguish an absent key from one explicitly assigned undefined. Removes redundant guards whose branches the surrounding NaN-fallback arithmetic already collapses to the identical result (sheetFormatDefaultRowHeightPt's raw-undefined check, readColumns' widthRaw-defined check, readRows' htRaw-undefined check), and the two early-return size checks in applyCellComments/applyCellResidueRules, whose absence only skips pointless work over an empty collection rather than changing any observable output. Documents two remaining genuinely irreducible equivalent mutants (the sqref-split regex's + quantifier, and fallbackEmptyWorksheet's own tag string, matching drawings.ts's identically-shaped case) with the exact reasoning that makes them unobservable through this module's own contract.
Removes applyCellResidueRules' own empty-firstToken disjunct:
parseRangeReference('') already returns undefined rather than
throwing, so the check was a redundant special case of the
undefined branch beside it, which stays load-bearing on its own.
Asserts displayText, not just value, for a true boolean cell,
closing the one real remaining gap in deriveDisplayText.
Documents deriveDisplayText's own "empty" case as a structurally
required but genuinely unreachable switch arm: ContentCellValue's
type still includes "empty" as a member, so the case must stay for
the function to type-check as returning string unconditionally,
even though neither of its two real call sites can ever pass one.
signature mutation gaps Adds direct unit coverage for every optional-field presence check in the read side (font/fill/border/alignment key absence, not just value, proven with hasOwn rather than toBeUndefined), for the exact val-string behaviour of readFontToggle/readFontUnderline, for a non-integer numFmtId/sizePt leaving the code/sizePt unresolvable, and for colorFromElement's own validation (invalid hex, a too-short rgb attribute). Adds write-side tests proving every one of CellFormatTable's own signature segments (each font/fill/border/alignment flag) actually distinguishes two otherwise-identical entries, rather than trusting that the signature strings alone don't collapse two different inputs onto the same interned index; and that a font/fill/border interned twice under different number formats still caches to a single declared entry. Removes readFontTableEntry's own redundant szVal-undefined guard: Number(undefined) is NaN, so an absent <sz val> already falls through the Number.isFinite check below to the same "no sizePt" result this guard would have selected directly. Documents two remaining genuinely irreducible equivalent mutants in colorFromElement (the raw.length >= 6 vs > 6 boundary, and the hex regex's own anchors) with the exact reasoning that makes them unobservable given hex's own fixed construction.
Adds a real regression test for the border style ?? "solid" fallback: an edge with no style stated and one explicitly styled "solid" now assert to the same borderId, proving the fallback actually merges two representations of the identical visible border rather than only avoiding a crash. Adds tests for fontFamily/sizePt/colour genuinely differing from or absent against the baseline, size/fontFamily alone distinguishing two interned fonts, an empty borders object not colliding with a real one, and internFill's own default branch throwing for a fill kind this discriminated union has no member for. Removes readBorderEdge's own redundant "none" special case: "none" is not a key XLSX_BORDER_STYLE declares, so it already falls through the resolved-undefined check below to the identical result this check would have returned directly. Documents the remaining genuinely irreducible equivalent mutants in the read-side and write-side non-integer-numFmtId guards (each redundant with a sibling guard on the only real call path) and in every internal-only, never-exposed signature-building segment (font/fill/border/alignment dedup keys), where no consistent relabelling or placeholder substitution can ever create a real collision given the actual domain of values each field carries.
actually observe their own mutant
Both replacement tests compared an empty-borders {} decoration
against a real one, but an empty {} decoration's own outer
signature already coincides with EMPTY_DECORATION's (the loop over
its zero edges appends nothing), so it hits the seeded
cellFormat-level cache before internFormat/internBorder is ever
called at all -- neither test could ever have observed a change to
internBorder's own per-edge signature building or to the outer
alignment-presence check, regardless of mutation.
Verified directly, per this project's own equivalence-claim
convention: applying each mutation by hand and running the affected
test confirmed it passed unchanged either way, before rewriting it.
The style ?? "solid" fallback now compares two SAME-number-format
interns (an implicit-style edge against an explicit "solid" one),
which genuinely forces reuse of the outer cellFormat cache and so
exercises signatureOfDecoration's own fallback, not
internBorder's separately-correct borderToXlsxStyle handling of
the same case.
The edge-segment test now compares two distinct REAL borders (both
of which genuinely reach internBorder) rather than an empty one
against a real one, so a collapsed per-edge segment is observable
as a wrongly-shared borderId.
Both cases were entirely unreached by any existing test -- every border test so far exercised only the dashed/solid weight-bucketing branches, leaving the two fixed-token cases genuinely uncovered rather than merely untested for a specific input.
This module had no dedicated test file at all: its only coverage came from content.test.ts/build.test.ts round trips through readXlsxContent, whose reader never inspects an OOXML element's exact tag or attribute spelling, only its structural shape -- so a round trip could never tell a real element name from a mutated one apart. Asserts the full xdr:oneCellAnchor/xdr:pic/xdr:graphicFrame shape for both a picture and a chart anchor, the drawing/chart namespace declarations, the relationship part's own Id/Type/Target triples, the chart XML declaration and c:chartSpace root, every c:ser/ c:barChart/c:catAx/c:valAx element and its fixed axis ids, the series/category range arithmetic against a non-trivial category count (so the +1 in each range's own upper bound, and the +1 from column index to letters, are both observable), object-id and relationship-id counters advancing across multiple images, media/ chart numbering advancing across two calls sharing one counters instance (the shared-across-sheets contract DrawingCounters' own doc comment states), and every one of this module's five thrown error paths (svg image, non-chart object, a missing anchor field, a non-spreadsheet document, a spreadsheet document with no sheet).
Moved the shared picture+chart buildSheetDrawing() call from the describe body into beforeEach. Stryker's per-test coverage instrumentation attributes a line's execution to whichever test is running when that line executes, and a describe body runs during test collection, before any it() has started -- a call made there is invisible to that attribution, so Stryker silently ran some other, less precise test against these mutants instead of this file's own assertions. Verified directly: an L62 StringLiteral mutant on "xdr:col" showed Survived in a real scoped run despite the exact assertion in this file catching it when the same mutation was applied by hand and run locally, until the call moved into beforeEach.
…arse-cell gaps Adds a two-chart test proving the object-id counter genuinely advances (2 then 3) for charts, not just for the two-image case already covered, and a sparse-chart-cells test proving a missing series-name/value cell reads back through chartSeriesFromDocument's own cellAt fallback as an empty string, matching the empty label a missing point already gets on the read side.
This module's only prior coverage came from two real-producer fixtures (cellIs, colorScale) and a round-trip suite covering every rule family's own value shape, neither of which exercised the module's own exact XML vocabulary directly: an attribute name mutation on the write side and the matching read-side lookup can cancel each other out in a round trip, and toEqual-based value checks cannot distinguish a key genuinely absent from one assigned undefined. Covers, on the read side: the wrapper-level sqref gate, priority/ stopIfTrue/source residue capture, every cfvo type token including formula/percentile (previously entirely uncovered), the colorScale cfvo/color count-mismatch rejection, dataBar/iconSet's own true-default showValue convention and reverse flag, every dxf residue passthrough branch (font/fill/numFmt/alignment/border/ protection, including a font or fill kept whole when no colour was captured from it), cellIs formula2 restricted to between/ notBetween, top10's rank<=0 rejection and percent/bottom presence, aboveAverage's own true-default and stdDev<=0 rejection, and the colorScale/iconSet type discriminants. Covers, on the write side: formula/formula2 element count, top10/ aboveAverage/dataBar/iconSet's own attribute presence (each written only when explicitly set, never restating a default), the colorScale element's cfvo-before-color ordering, and buildConditionalFormattingElements' own range-based grouping and gap-filling priority assignment (an unpriorised rule never reuses an already-claimed explicit priority).
…sidue gaps Covers every isSheetRuleOperator member distinctly (not just between/greaterThan), cellIs formula2 for notBetween alongside between, the absent-timePeriod rejection, the colorScale cfvo/ color count boundary at exactly 2 and exactly 3 stops on both sides, residualAttributesFor's own expectedTag gate (an unmanaged cfRule attribute genuinely restored on write), rangeSetKey's field separator (two ranges that would collide under naive concatenation without it), and a full round trip of every dxf residue kind at once (font+color, fill+patternFill+bgColor, numFmt, alignment, border, protection) through DxfTable.intern. Adds the missing "omitted entirely" half of several write-side attribute-presence tests (top10 percent, aboveAverage/equalAverage/ stdDev, dataBar showValue, iconSet reverse/showValue) that only asserted the explicit-true/false case, never that the attribute is genuinely absent -- not merely unasserted -- when nothing was set.
… underline/strike tokens Adds a minimal, layout/master-free slide package builder for isolating a single shape's own paragraph/run properties, and uses it to cover: the widescreen-default fallback when p:sldSz carries no cx (previously untested against a real, differently-sized sldSz, so the default and a genuine explicit size were indistinguishable), every algn token (l/ctr/r/just/justLow, plus an unrecognised token falling through to no alignment), and the exact none/noStrike tokens for underline and strikethrough alongside their positive and absent-attribute cases.
re-measured floor Re-measures the package-wide score after closing the mutation gaps in content.ts, styles.ts, drawings-write.ts (now a genuine 100%), conditional-format.ts, and pptx/read.ts: 84.44% of 6891 valid mutants, up from the original 64.55% baseline the threshold of 63 reflected. States plainly which modules remain the real next targets (xlsx/build.ts, docx/write.ts, docx/read.ts), so the number reads as a measured floor rather than a ceiling.
Mearman
force-pushed
the
feat/100-percent-mutation-ooxml.js
branch
from
September 16, 2026 05:53
856c7dc to
e59d726
Compare
Adds direct assertions for the XML declaration prolog, every [Content_Types].xml Override across a document exercising comments, images, charts, and tables, the fixed _rels/.rels and xl/_rels/workbook.xml.rels relationships, xl/workbook.xml's sheetId and r:id numbering, and xl/sharedStrings.xml's count/uniqueCount and xml:space attribute. Closes the Print_Titles derivation gap where repeatRows and repeatColumns were always set together, so mutating the || between them to && never changed the observable output; adds cases with each set alone and with neither set.
Asserts the exact fixed scaffolding buildStylesPart writes for a document needing no number formats: the single default font, the two reserved fills, the one reserved border, and the default cellStyleXfs/cellXfs/cellStyles entries, none apply*-flagged. Adds a font carrying bold, italic, strike, and underline together, proving each toggle writes its own element independently; a border with only its top edge set, proving the per-edge branch runs independently for each of the four edges rather than uniformly; a cell with verticalAlignment 'top', the one branch neither 'middle' nor the default omission exercises; and pattern fills with only a foreground or only a background colour. Asserts every docProps/core.xml and docProps/app.xml field, including subject, modifiedIso, and creator, and the case where metadata carries none of them and keywords is an empty array.
…and merge output exactly Adds cases where a sheet's dimension is extended solely by its columns array or solely by its rows array (no cells at all), and a case where a column/row entry reaches past the last cell, proving computeDimension takes the genuine max across all three sources rather than letting one silently dominate. Asserts buildColsElement writes hidden with no width attribute, and width/customWidth with no hidden attribute, as two independent column declarations rather than always pairing the two. Proves buildSheetDataElement sorts both rows and, within a row, cells into ascending order regardless of input order, and that a row with no matching ContentSheetRow entry carries only its own r attribute. Proves buildMergeCellsElement treats colSpan and rowSpan as independent merge triggers, and writes no <mergeCells> at all when every cell's span is 1 or absent. Covers buildCellElement's decoration ternary for alignment-only cells, the exact <f>/<v> child order and missing t attribute for a formula's numeric result, and the no-formula case writing no <f> at all; renderString's formula-result branch for an unparseable temporal value; and buildSheetPrElement's fitToPage reflecting whether the sheet actually declares fitToPages.
… breaks exactly Asserts ptToInches's genuine points-to-inches conversion for both the common 72pt case and non-72pt margins, and the fixed 0.3in header/footer margin. Covers buildPageSetupElement's paperSize/paperWidth-paperHeight branch for a standard versus a custom page size, the landscape and portrait orientation branches, the default scale/fitToWidth/ fitToHeight when neither scalePercent nor fitToPages is declared, and their declared values when present. Covers buildBreaksElements writing row breaks and column breaks independently of each other, with the exact id/min/max/man attributes, and writing neither element when manualBreaks is undefined or both its arrays are empty. Asserts buildWorksheetPart writes cols/mergeCells/drawing/ tableParts together for a sheet carrying every optional feature and none of them for a plain sheet, plus the worksheet root's own xmlns/xmlns:r and buildWorksheetRelsPart's Relationships root.
…onship numbering exactly Proves a table definitions entry attaches tableParts and its own xl/tables/tableN.xml only to the sheet it names, and that an unrelated second sheet gets neither the table nor its own rels part at all, closing the gap where the 'table.sheet !== sheet.name' skip was never exercised by a genuinely non-matching sheet. Asserts worksheet relationship ids are assigned sequentially (rId1/rId2/rId3) across comments, drawing, and table relationships on the same sheet, in that order. Proves usedImageFormats collects every distinct format a sheet's images actually use (png and jpeg together), not just the first, and declares no Default extension for a format never used. Adds a negative assertion that a plain document with neither a chart nor a table writes no /xl/charts/ or /xl/tables/ Override at all, closing the gap where an initial-empty-array mutant seeding a bogus part name went undetected by toContainEqual-only assertions.
…content The existing "derives the reserved print names" test checked the definedName's name and localSheetId but never its own text, so a mutant dropping the range text entirely went unnoticed.
…ryker reporting anomaly xlsx/build.ts now measures 69.3-69.5% of its own valid mutants under Stryker's scoped mutation run, up from 32.83% before the structural coverage added in this session's earlier commits, confirmed reproducible across three independent runs including one at concurrency 1. Documents a genuine tool-measurement anomaly found while verifying that improvement: several mutants Stryker's own reporter marks survived were directly disproven as equivalent by manually applying the exact mutation and running the identical vitest configuration Stryker's own runner uses, which fails the relevant tests every time. The package breakThreshold is left unchanged, since raising it needs a fresh full-package measurement once docx/write.ts and docx/read.ts, the package's two remaining large modules, have also been closed.
Narrow worksheetRels[0] with an explicit undefined check before passing it to attr(), instead of an unsound index access typed as XmlElement | undefined, and drop the readonly modifier from pkgWithBreaks's manualBreaks parameter to match ContentSheetPrintSettingsSchema's own mutable array type, which readonly arrays cannot satisfy.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Part of the workspace-wide effort to bring every package's Stryker mutation score to a genuine 100% with zero disable comments (see the sibling PRs already merged for byte-codec, excel-number-format, document-compute.js, pdf-raster-cpu).
Measured baseline for ooxml.js: 65.05% of 7126 valid mutants (killed 4604, timeout 31, survived 2081, no-coverage 410).
This PR is a work in progress. It has landed, in order:
test-support/cfb.ts), including byte-level assertions for header fields archive-codec's own reader deliberately never cross-checks, and a fixture sized to the builder's own one-FAT-sector boundary.xlsx/defined-names.tsandxlsx/data-validation.tsbrought to 100% from zero direct coverage, plus a handful of redundant-guard removals proven equivalent by hand-mutation (a name-then-type check that already excluded an absent name, a no-op ternary branch, an indexOf/slice split removing a possibly-undefined array index, a letters regex made redundant by columnLettersToIndex's own character-class check, and a print-area builder rewritten to construct its dollared reference directly from row/column indices instead of formatting then re-parsing a plain reference through a regex).xlsx/print-settings.tsbrought to 100%: margin/break/scale coverage plus one redundant scale-presence guard removed (Number(undefined) is NaN, already caught by the isFinite check beside it).docx/numbering.ts,pptx/chart.ts, andpptx/diagram.tsbrought to 100%: direct unit coverage for readChartTable/readChartResidue and readDiagramText/readDiagramResidue (previously exercised only indirectly through full xlsx/pptx round trips), plus the numbering override's no-w:val case.docx/styles.tsbrought to 100%: style-cascade type discrimination (a same-styleId style of the wrong type, a default-style flag on the wrong type), the default paragraph style's own w:pPr merge, strike's basedOn inheritance, majorAscii/minorAscii alongside their HAnsi spellings, an unrecognised asciiTheme resolving to no font, a bare w:u with no w:val, w:ind/@w:start as w:left's fallback, "distribute"/"atLeast" alongside their siblings, a themeTint byte with stray characters before or after its two hex digits, plus one redundant guard removed (readToggle's absent-value check, subsumed by the three inequality comparisons beside it).pptx/reading-order.tsfrom 74.63% to 89.47%: an exact axis-ratio tie breaking to rows, overlapping shapes whose primary/secondary sort keys disagree, a row needing its own internal column cut once split out, extentAlong's true-span computation, and two provably-redundant guards removed (a shapes.length<=1 early return the algorithm's own fallback already makes a no-op, and a columns.groups.length>1 check implied by the ratio comparison already being positive). Three touching-boundary mutants in splitOnGap remain genuinely difficult to distinguish through the exported function's output alone and are left for a follow-up pass.Current measured score: 72.05% of the package's valid mutants (up from the 68.17% this PR previously reported, and the 65.05% original baseline). Zero Stryker disable comments anywhere in the package (
grep -rn "Stryker disable" src/returns no matches).Files now at a genuine 100% mutation score in this PR:
util/base64.ts,xml/build.ts,xml/parse.ts,model/node.ts,typed/util.ts,typed/xlsx/util.ts,typed/xlsx/defined-names.ts,typed/xlsx/data-validation.ts,typed/xlsx/print-settings.ts,typed/docx/numbering.ts,typed/docx/shading.ts,typed/docx/styles.ts,typed/pptx/chart.ts,typed/pptx/diagram.ts,typed/pptx/inherit.ts,typed/xlsx/serial.ts,typed/xlsx/definitions.ts,typed/xlsx/rule-residue.ts,typed/xlsx/comments-write.ts,typed/xlsx/shared-strings.ts,typed/xlsx/sqref.ts,typed/xlsx/units.ts,typed/shared/units.ts,typed/embedded.ts,typed/document-tree.ts,typed/figure-captions.ts, plusimage/sniff.tsandpackage-io/read.ts/write.ts.Every fix is either a real test proving a genuine behavioural difference, or a small refactor removing code whose mutation was verified equivalent by hand (mutating the source directly and confirming the existing suite still passed before the fix, and failed after it) -- never a Stryker disable comment.
Substantial work remains: the package's largest modules are still far below 100% --
typed/docx/write.ts(57.81%, 2124 lines),typed/docx/read.ts(75.98%, 1896 lines),typed/xlsx/build.ts(33.58%, 1235 lines),typed/pptx/read.ts(77.23%, 1048 lines),typed/xlsx/conditional-format.ts(62.09%),typed/xlsx/styles.ts(68.83%),typed/xlsx/content.ts(57.00%),typed/xlsx/drawings-write.ts(40.82%),typed/xlsx/drawings.ts(65.13%),typed/xlsx/comments.ts(75.00%),typed/docx/constructs.ts(80.77%),typed/pptx/reading-order.ts(89.47%, close),typed/shared/color.ts(66.42%),typed/shared/drawingml.ts(76.88%),typed/shared/metadata.ts(76.00%),typed/compact.ts(47.95%),typed/xlsx.ts(60.00%),typed/xlsx/definitions-write.ts(54.00%), andtest-support/embedded.ts(65.00%).stryker.config.ts'sbreakThresholdis left at its original measured-baseline value (63) rather than raised, since the package is not close to the genuine 100% that value would need to reflect.Left as a draft while this continues.