Fix ReplaceSpecialCharacters() silently dropping its dot-collapse step

[vertesy]

The bug

The function computed ReplaceRepeatedDots(x) as its second-to-last statement but never assigned the result back to x — it was a bare expression, so R discarded it. The actual return value came from the final if (remove_dots) x <- gsub(...) else x line, which still operated on the un-collapsed x.

Impact

A direct call to ReplaceSpecialCharacters() on a string with adjacent special characters returns runs of repeated dots instead of a single collapsed dot, e.g. ReplaceSpecialCharacters("a[[b]]") returned "a..b.." instead of "a.b." (verified before/after). Most existing call sites (FixPlotName(), ParseFullFilePath()) were unaffected in practice, because they each independently call ReplaceRepeatedDots() again on the result afterward and so already compensated for this — but any code calling ReplaceSpecialCharacters() directly got the uncollapsed output.

Fix

Assign the result (x <- ReplaceRepeatedDots(x)) so the collapse step actually applies before the optional remove_dots step.

Version bumped 1.2.1 → 1.2.2 in Development/config.R and DESCRIPTION.

Validation

Manually confirmed direct calls now collapse dots correctly while FixPlotName()/ParseFullFilePath() output is unchanged (as expected, since they already compensated). R CMD check and the existing testthat suite (7/7 passing) were run on this branch. Remaining R CMD check findings (parFlags example error, parsepvalue undocumented prefix, %!in% Rd-name warning, sandbox locale warning) are pre-existing and out of scope, addressed in other PRs in this batch or environment-inherent.

---

Generated by Claude Code