Cross-workspace search is complete only when both dimensions agree: how many matches exist and how many files contain them. The controlled procedure uses eight chapter files, fourteen expected matches distributed across six chapters, and two excluded reference files under notes and archive. Fourteen and six are reader-run expectations, not values measured in Rune Studio during hands-on verification. This guide locates an omission in query, scope, or manuscript version. It does not edit or replace any match.
Keep two expectations: fourteen matches and six files
The chapter distribution is 3, 2, 0, 4, 1, 0, 2, and 2, totaling fourteen. Chapters 1, 2, 4, 5, 7, and 8 contain matches, so the file count is six.
A total alone hides a match moving from one chapter to another. A file count alone hides a missing occurrence inside the same six files. Both values must agree.
Exclude notes and archive by scope
Search only the eight files under chapters. The notes and archive files contain five additional occurrences but are outside the current-manuscript question. Confirm the selected folder path before searching.
Do not delete excluded occurrences merely to make the number smaller. Their documented exclusion explains a nineteen-match whole-workspace result.
Record the first result as a reproducible baseline
Write query, selected folder, manuscript version, time, match total, and matching-file total. “Term A / chapters / v12 / 13 matches in 5 files” is useful; “search was short” is not. Append each changed condition as a new row.
Use 13 matches in 5 files to find a missing chapter
When both totals are lower by one, first compare the file list with the six expected chapter names. If chapter 5 is absent, check its selection, extension, unsaved state, and version.
If chapter 5 is present, compare per-file counts because another chapter may contain the missing line. Equal numerical differences do not prove one cause.
Treat 14 matches in 5 files as a distribution failure
A correct match total with one missing file can mean that a chapter disappeared while another gained a match. Compare all eight per-file values, including chapters expected to have zero. Total agreement is not completion.
Use 15 matches in 6 files to locate one extra line
The six target files are present, so use per-file and per-line grouping to locate the additional occurrence. Open the result in context, but do not delete it in this workflow. Hand its file and line to a semantic review.
Recognize 19 matches in 8 files as scope expansion
The two excluded files add five matches and two files. This pattern points to whole-workspace scope rather than a bad query. Restrict the target to chapters and check whether fourteen and six return.
Use the counts and scoped search verified in Rune Studio
In hands-on verification, a four-file specimen contained four matches. Limiting the target to text files in the body folder produced three matches in two files, while one match in an excluded reference file remained outside the scope. The search output listed results by file and line, allowing the match count and matching files to be compared.
Real-time interface updates, clicking a result, and this guide’s fourteen-in-six example were not exercised. The confirmed count, scope, file, and line records support the method without stating that the application interprets meaning or repairs omissions automatically.
Account for case-insensitive English search
Documented workspace search does not distinguish English uppercase and lowercase. If API, Api, and api must be separate categories, do not reuse one combined count. Define an expectation that matches the search behavior or use a separate structure check.
Freeze the eight-row expectation before search
List chapter01 through chapter08 with 3, 2, 0, 4, 1, 0, 2, and 2 before seeing the result. Keep the two zero rows because they reveal an unexpected match. Do not rewrite the expectation to fit the observed output.
Read the verified file and line records in two stages
First compare the six expected file names, then check per-line occurrences inside each. Record whether a result is in prose, heading, or note. The hands-on check confirmed file-and-line output, but not opening results by clicking them in the interface. The workflow locates the difference; it does not draft a correction.
Keep unsaved-tab and UI behavior explicitly unverified
Documentation describes unsaved-tab handling for workspace replacement, but this search check did not exercise unsaved-tab count behavior. If observed counts differ from the v12 expectation, treat open unsaved tabs as a condition to check. Either save and update the expected version or exclude the unsaved change. Do not combine two source states to force fourteen.
Convert each numerical pattern into a return point
Use 13/5 for a missing file, 14/5 for distribution, 15/6 for an extra line, and 19/8 for expanded scope. Hand the expected-versus-observed pair forward instead of a vague complaint about search.
Prove that zero-match chapters were in scope
List all eight targets and mark chapters 3 and 6 as included with expected zero. Six matching files does not mean only six files were searched. Keep target-file count separate from matching-file count.
Verify extension and manuscript version beside each name
Distinguish chapter05.md v12 from chapter05-old.md or a text duplicate. Compare result paths, not only base names, so an archive chapter is not counted as current.
Preserve fourteen short result excerpts
Store a short excerpt for each numbered occurrence. If one line contains two matches, assign two occurrence numbers. The list explains how six files produce fourteen matches without completing semantic review.
Finish by explaining fourteen and six
For a reader who creates this specimen, completion is the distribution 3, 2, 0, 4, 1, 0, 2, 2, fourteen total matches, and six matching chapter files. This is the article’s expected result, not a completed hands-on verification run. No occurrence is edited.
Select chapters and review the search-count and scope behavior on the Rune Studio product page. Hand single-file context to the single-file context-classification workflow and replacement to the bounded multi-file replacement workflow or the single-chapter exception-aware replacement workflow.