- Options automatically fallback to factory defaults if they are unset (h/t @gadenbui).
- #158: Calling
diff*
withdo.call
now works without warnings. - #117: Fix guide detection with very wide wrapped data.frames (h/t @bastician, @overvolting).
- Add a print method for
ses_dat
return values that makes it easier to interpret the diff. - #152: Rewrite the
fall-back "O(n)" algorithm that kicks in when there are
max.diffs
differences to be more robust (h/t @hadley, @DanChaltiel, @gadenbui). - Related to #152:
max.diffs=0
used to mean the same asmax.diffs=-1
(i.e. unlimited), but this was undocumented and an error.max.diffs=0
will now immediately fall back to the "O(n)" algorithm.
- Implement experimental .Rout / .Rout.save testing.
- Fix
all.equal
test breakages from r79555.
- #150: Make tests compatible
with new
testthat
release (h/t @hadley). - Remove pre-built vignettes and note
testthat
change towaldo
release.
- #143: Add
ses_dat
to provide a more computable version ofses
(h/t @hadley). - #144: Re-encode strings to UTF-8 prior to comparison to avoid spurious encoding-only differences (h/t @hadley).
- #142: Typos in
standardGeneric
in trim/guide generic definitions. - Drop attributes from inputs to
diffChr
(revealed as an issue by #142). - Banish ghosts of
stringsAsFactors
.
- Tests explicitly set
stringsAsFactors=TRUE
so they don't fail with the anticipated changed for R4.0. - #140: Bad link in
?ses
.
This is a bugfix release.
- #136: Documentation for
ignore.white.space
(h/t @flying-sheep) andmax.diffs
parameters listed incorrect defaults. - #135: Incorrect handling of potential meta data strings when unwrapping atomics would cause a "wrong sign in by argument" error (h/t @flying-sheep). We also fixed other bugs related to the handling of meta data in atomic vectors that were uncovered while debugging this issue.
- #134: Forwarding
...
todiff*
functions no longer breaks substitution of arguments for diff banners (h/t @noamross).. - #133:
diffFile
considers files with equal content but different locations to beall.equal
now (h/t @noamross). - #132: Duplicate pager slot
for baseline
Pager
removed (h/t Bill Dunlap).
There are also several other small internal changes that in theory should not affect user facing behavior.
- Set
RNGversion()
due to changes to sampling mechanism.
- #129: Allow pager
specification via lists rather than full
Pager
objects for easier changes to defaults. As part of this we changedStyleRaw
objects to use default pager instead ofPagerOff
. - #126: Add embedding diffs in Shiny to vignette.
- #119:
ignore.whitespace
now also ignores white space differences adjoining punctuation. - #118: New option to preserve
temporary diff file output when using pager (see
?Pager
). - #114: New options
strip.sgr
andsgr.supported
allow finer control of what happens when input already contains ANSI CSI SGR and how ANSI CSI SGR is handled in string manipulations. Related to this,options(crayon.enabled=TRUE)
is no longer set when capturing output prior to diff as it used to be. By default pre-existing ANSI CSI SGR is stripped with a warning prior to comparison.
- #131: Fix missing slot in S4 class definition (discovered by Bill Dunlap).
- #127: Width CSS conflicts with bootstrap (reported by @eckyu, debugged/fixed by @cpsievert).
- #123: Compatibility with R3.1 (@krlmlr).
- #121: Vignette describing how to embed diffs in Rmd documents (@JBGruber).
- #115: Declare HTML page diff encoding/charset as UTF-8 (@artemklevtsov).
- Comply with CRAN directive to remove references to packages not in depends/imports/suggests in tests (these were run optionally before).
- Fix bugs in corner case handling when we hit
max.diffs
.
- Fix test failures caused by changes in tibble output
- #111: Fixed guides with
row.names=FALSE
(thank you @Henrik Bengtsson). - #113: Adapt tests to new
str
return values (thank you @Martin Mächler).
- Fix tests for next
testthat
release. - #107: Diffs on quoted language
- #108: Problems caused by
copying
crayon
functions (@seulki-choi, @gaborcsardi) - #100: R_useDynamicSymbols
- #97: 2D Guidelines fixes for data.table, tibble
- #96: Warnings when comparing large data tables.
- #94: Guide detection problems in nested lists.
- #105: Copyright tweaks.
- #87:
diffobj
is now GPL (>=2) instead of GPL-3. - #81: Better handling of mixed UTF-8 / ASCII strings, reported by jennybc
- #88: correctly handle trimming when empty lists are involved, reported by wch
- #77:
diffObj
now favors dispatching todiffPrint
unlessdiffPrint
output is large - #82:
diffChr
andses
now treatNA
as "NA" (needed with change innchar(NA)
in base R) - #85: Improved alignment of unwrapped atomic vector diffs
- #83: Improve pager auto
detection (note now ANSI output is only allowed by default if terminal
supports ANSI colors and the system pager is
less
, see?Pager
for details) - #92, #80, #45: basic implementation of S4 guidelines and trimming (full resolution eventually with #33)
- #84: simplify how to call
diffChr
for improved performance, including "optimization" ofconvert.hz.whitespace
. - #64: fix line limit in corner case
- More robust handling of external
diff*
methods and of howdiffObj
callsdiffStr
anddiffPrint
- #71: Buggy diffs b/w data frames when one has sequential row numbers and the other does not, loosely related to #38
- #69: Improve performance on outputs with large print/show output, and other assorted minor optimizations
- #72: Warn when
style
parameter overrides other user supplied parameters - #70: Improve word contrast in YB HTML mode
- #63: Show
all.equal
output when objects are notall.equal
but there are no visible differences - Add Mean Relative Indifference vignette and update vignette styling
- #67: Fix CRAN Binaries
- Clarified that C code is heavily modified and incompatible with original
libmba
implementation
- First version on CRAN
- #51: use RStudio viewport to display HTML diffs when running in RStudio (h/t Noam Ross)
- #54: #55, scale HTML output to viewport width (see
?Style
) - #53: default term colors computed on run instead of on package load
- #56: disable wrap for HTML output
- HTML output now captured with default width 80 since there is no explicit relationship between HTML viewport width and
getOption("width")
- The
style
parameter now accepts lists to use as instantiation arguments forStyle
objects (see?Style
) - Fix subtle rendering and formatting application flaws
- Switch Travis shields to SVG per Gábor Csárdi
- Improve in-hunk alignment of partially matching lines
- Compile with
-pedantic
, fix related warnings Arun - Improved coverage and more robust testing
- Several internal structure changes to accommodate improvements
- #46: Guide and Trim Problems with Lists
- #47: Output Format in non-ANSI Terminals Without Browser (reported by Frank)
- #48:
make_blocking
Default prompt Confusing (reported by Frank) - #49: In-Hunk Word Diffs Issues when Unwrap-diffing Atomics
- #50: CSS Lost in Rstudio Server Sessions (reported by Steven Beaupré)
- Turn off unwrapping for named atomic vectors (see #43)
- #44: Proper handling of NULL objects in
diffStr
- #41: Compilation Issues in Winbuilder
- Initial Release