Simplified Array::wrap2 algorithm and performance improvement

performance improvement using: - echo instead of string concatenation - string substitution instead of calling sed on each element - echo -e removed the need to do a loop on each character to parse ansi code and the need of Filters::removeAnsiCodes
fchastanet · Jan 13, 2024 · bb58e72 · bb58e72
1 parent 30f45db
commit bb58e72
Show file tree

Hide file tree

Showing 33 changed files with 1,409 additions and 1,616 deletions.
diff --git a/.github/workflows/lint-test.yml b/.github/workflows/lint-test.yml
@@ -11,6 +11,9 @@ on: # yamllint disable-line rule:truthy
 jobs:
   build:
     runs-on: ubuntu-22.04
+    permissions:
+      # needed by ouzi-dev/commit-status-updater@v2
+      statuses: write
     strategy:
       fail-fast: true
       matrix:
@@ -111,7 +114,7 @@ jobs:
       - name: Archive results
         if: matrix.runPrecommitTests && always()
         continue-on-error: true
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: linter-reports
           path: |
@@ -151,7 +154,7 @@ jobs:
 
       - name: Upload Test Results
         if: always()
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: Test Results ${{matrix.vendor}} ${{matrix.bashTarVersion}}
           path: logs/report.xml

diff --git a/.markdown-link-check.json b/.markdown-link-check.json
@@ -6,7 +6,10 @@
       "pattern": "^http://localhost:3000/"
     },
     {
-      "pattern": "^https://www.cyberciti.biz/.*$"
+      "pattern": "^https://www.cyberciti.biz/"
+    },
+    {
+      "pattern": "^https://tinyurl.com/"
     },
     {
       "pattern": "^#.*"

diff --git a/.mega-linter.yml b/.mega-linter.yml
@@ -71,7 +71,8 @@ EDITORCONFIG_EDITORCONFIG_CHECKER_FILTER_REGEX_EXCLUDE: |
     bin/bash-tpl|
     ^doc/guides/Options/generate.*\.md$|
     ^pages/Commands.md
-    ^.*-megalinter_file_names_cspell.txt
+    ^.*-megalinter_file_names_cspell.txt|
+    testsData/.*\.result$
   )
 
 GIT_GIT_DIFF_PRE_COMMANDS:

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -11,7 +11,8 @@ repos:
         exclude: |
           (?x)(
             ^doc/images/.*\.svg$|
-            ^doc/guides/Options/generate.*.md$
+            ^doc/guides/Options/generate.*.md$|
+            testsData/.*\.result$
           )
       - id: check-executables-have-shebangs
       - id: check-shebang-scripts-are-executable
@@ -49,7 +50,11 @@ repos:
     rev: v4.0.0-alpha.8
     hooks:
       - id: prettier
-        exclude: ^doc/guides/Options/generate.*.md$
+        exclude: |
+          (?x)(
+            ^doc/guides/Options/generate.*.md$|
+            ^\.vscode/.*\.code-snippets$
+          )
 
   - repo: https://github.com/fchastanet/bash-tools-framework
     rev: 2.1.0

diff --git a/.vscode/bats.code-snippets b/.vscode/bats.code-snippets
@@ -15,6 +15,6 @@
       "[[ \"\\${status}\" = \"0\" ]]",
       "run cat \"\\${BATS_TEST_TMPDIR}/result\""
     ],
-    "description": "bats get status and output"
+    "description": "bats get status and output",
   }
 }
diff --git a/.vscode/shdoc.code-snippets b/.vscode/shdoc.code-snippets
@@ -24,6 +24,6 @@
       "# @require Git::requireGitCommand",
       "# @feature Retry::default, sudo"
     ],
-    "description": "bash function generic shdoc"
+    "description": "bash function generic shdoc",
   }
 }
diff --git a/README.md b/README.md
@@ -389,10 +389,18 @@ bin/compile src/_binaries/testsData/bin/embed.sh --template-dir src --bin-dir bi
 echo $? # prints 127
 
 # try to get more logs
-KEEP_TEMP_FILES=1 BASH_FRAMEWORK_DISPLAY_LEVEL=4 bin/compile src/_binaries/testsData/bin/embed.sh --template-dir src --bin-dir bin --root-dir $PWD --src-dir src/_binaries/testsData/src
+KEEP_TEMP_FILES=1 BASH_FRAMEWORK_DISPLAY_LEVEL=4 bin/compile \
+  src/_binaries/testsData/bin/embed.sh \
+  --template-dir src \
+  --bin-dir bin \
+  --root-dir "${PWD}" \
+  --src-dir src/_binaries/testsData/src
 
 # try to use strace
-docker run --rm -it -w /bash -v "$(pwd):/bash" --entrypoint="" build:bash-tools-alpine-4.4-user bash
+docker run --rm -it \
+  -w /bash -v "$(pwd):/bash" \
+  --entrypoint="" \
+  build:bash-tools-alpine-4.4-user bash
 apk update
 apk add strace
 ```

diff --git a/bin/awkLint b/bin/awkLint
@@ -101,10 +101,12 @@ trap cleanOnExit EXIT HUP QUIT ABRT TERM
 # @description concat each element of an array with a separator
 # but wrapping text when line length is more than provided argument
 # The algorithm will try not to cut the array element if it can.
-# if an arg can be placed on current line it will be,
+# - if an arg can be placed on current line it will be,
 #   otherwise current line is printed and arg is added to the new
 #   current line
-# empty arg is interpreted as a new line.
+# - Empty arg is interpreted as a new line.
+# - Add \r to arg in order to force break line and avoid following
+#   arg to be concatenated with current arg.
 #
 # @arg $1 glue:String
 # @arg $2 maxLineLength:int
@@ -126,117 +128,102 @@ Array::wrap2() {
     return 0
   fi
 
-  eatNextSpaces() {
-    if [[ "${currentChar}" != [[:space:]] ]]; then
-      ((i--)) || true
-      return 0
-    fi
-    for (( ; i < textLength; i++)); do
-      if [[ "${text:${i}:1}" != [[:space:]] ]]; then
-        ((i--)) || true
-        break
-      fi
-    done
-  }
   printCurrentLine() {
-    echo -e "${currentLine}" | sed -E -e 's/[[:blank:]]*$//'
+    if ((isNewline == 0)) || ((previousLineEmpty == 1)); then
+      echo
+    fi
     ((isNewline = 1))
-    currentLine="${indentStr}"
+    echo -en "${indentStr}"
     ((currentLineLength = indentNextLine)) || true
   }
-  nextLine() {
-    printCurrentLine
-    eatNextSpaces
-  }
-  local currentLine currentChar ansiCode
-  local -i isAnsiCode currentLineLength=0 isNewline=1 textLength=0
-  local text=""
-  local arg=""
-  while (($# > 0)); do
-    arg="$1"
-    shift || true
-    if ((${#arg} == 0)) || [[ "${arg}" = $'\n' ]]; then
-      if ((isNewline == 0)); then
-        # isNewline = 0 means currentLine is not empty
-        printCurrentLine
-      fi
-      echo
-      continue
-    fi
-    local textFirstLine="${arg%%$'\n'*}"
-    text="$(echo "${arg}" | sed -E '1d')"
-    ((textLength = ${#text})) || true
-    local textFirstLineNoAnsi
-    textFirstLineNoAnsi="$(echo "${textFirstLine}" | Filters::removeAnsiCodes)"
-    local -i textFirstLineNoAnsiLength=0
-    ((textFirstLineNoAnsiLength = ${#textFirstLineNoAnsi})) || true
-
-    if ((isNewline == 0)); then
-      glueLength="${#glue}"
+  appendToCurrentLine() {
+    local text="$1"
+    local -i length=$2
+    ((currentLineLength += length)) || true
+    ((isNewline = 0)) || true
+    if [[ "${text: -1}" = $'\r' ]]; then
+      text="${text:0:-1}"
+      echo -en "${text%%+([[:blank:]])}"
+      printCurrentLine
     else
-      glueLength="0"
+      echo -en "${text%%+([[:blank:]])}"
     fi
-    if ((currentLineLength + textFirstLineNoAnsiLength + glueLength > maxLineLength)); then
-      if ((isNewline == 0)); then
-        # isNewline = 0 means currentLine is not empty
+  }
+
+  (
+    local currentLine
+    local -i currentLineLength=0 isNewline=1 argLength=0
+    local -a additionalLines
+    local -i previousLineEmpty=0
+    local arg=""
+
+    while (($# > 0)); do
+      arg="$1"
+      shift || true
+
+      # replace tab by 2 spaces
+      arg="${arg//$'\t'/  }"
+      # remove trailing spaces
+      arg="${arg%[[:blank:]]}"
+      if [[ "${arg}" = $'\n' || -z "${arg}" ]]; then
         printCurrentLine
+        ((previousLineEmpty = 1))
+        continue
+      else
+        if ((previousLineEmpty == 1)); then
+          printCurrentLine
+        fi
+        ((previousLineEmpty = 0)) || true
       fi
-      # restore current arg without considering first line
-      text="${arg}"
-      ((textLength = ${#text})) || true
-    else
-      if ((isNewline == 0)); then
-        # isNewline = 0 means currentLine is not empty
-        currentLine+="${glue}"
-        ((currentLineLength += glueLength)) || true
+      # convert eol to args
+      mapfile -t additionalLines <<<"${arg}"
+      if ((${#additionalLines[@]} > 1)); then
+        set -- "${additionalLines[@]}" "$@"
+        continue
       fi
-      currentLine+="${textFirstLine}"
-      isNewline="0"
-      ((currentLineLength += ${#textFirstLine})) || true
-    fi
 
-    for ((i = 0; i < textLength; i++)); do
-      currentChar="${text:${i}:1}"
-
-      if [[ "${currentChar}" = $'\r' ]]; then
-        # ignore
-        true
-      elif [[ "${currentChar}" = "\x1b" ]]; then
-        isAnsiCode=1
-        ansiCode+="${currentChar}"
-      elif ((isAnsiCode == 1)); then
-        ansiCode+="${currentChar}"
-        if [[ "${currentChar}" =~ [mGKHF] ]]; then
-          isAnsiCode=0
-          echo -e "${ansiCode}"
-        elif [[ "${currentChar}" = $'\n' ]]; then
-          # invalid ansi code, ignore it
-          isAnsiCode=0
-          ansiCode=""
+      ((argLength = ${#arg})) || true
+
+      # empty arg
+      if ((argLength == 0)); then
+        if ((isNewline == 0)); then
+          # isNewline = 0 means currentLine is not empty
+          printCurrentLine
         fi
+        continue
+      fi
+
+      if ((isNewline == 0)); then
+        glueLength="${#glue}"
       else
-        # non ansi code
-        if [[ "${currentChar}" = $'\n' ]] || ((currentLineLength == maxLineLength)); then
-          nextLine
-        elif [[ "${currentChar}" = "\t" ]]; then
-          if ((currentLineLength + 2 <= maxLineLength)); then
-            currentLine+="  "
-            ((isNewline = 0)) || true
-            ((currentLineLength = currentLineLength + 2))
-          else
-            nextLine
-          fi
+        glueLength="0"
+      fi
+      if ((currentLineLength + argLength + glueLength > maxLineLength)); then
+        if ((argLength + glueLength > maxLineLength)); then
+          # arg is too long to even fit on one line
+          # we have to split the arg on current and next line
+          local -i remainingLineLength
+          ((remainingLineLength = maxLineLength - currentLineLength - glueLength))
+          appendToCurrentLine "${glue:0:${glueLength}}${arg:0:${remainingLineLength}}" "$((glueLength + remainingLineLength))"
+          printCurrentLine
+          arg="${arg:${remainingLineLength}}"
+          # remove leading spaces
+          arg="${arg##[[:blank:]]}"
+
+          set -- "${arg}" "$@"
         else
-          currentLine+="${currentChar}"
-          ((isNewline = 0)) || true
-          ((++currentLineLength))
+          # the arg can fit on next line
+          printCurrentLine
+          appendToCurrentLine "${arg}" "${argLength}"
         fi
+      else
+        appendToCurrentLine "${glue:0:${glueLength}}${arg}" "$((glueLength + argLength))"
       fi
     done
-  done
-  if [[ "${currentLine}" != "" ]] && [[ ! "${currentLine}" =~ ^[\ \t]+$ ]]; then
-    printCurrentLine
-  fi
+    if [[ "${currentLine}" != "" ]] && [[ ! "${currentLine}" =~ ^[\ \t]+$ ]]; then
+      printCurrentLine
+    fi
+  ) | sed -E -e 's/[[:blank:]]+$//'
 }
 
 # @description ensure env files are loaded
@@ -523,19 +510,6 @@ Env::pathPrepend() {
   done
 }
 
-# @description remove ansi codes from input or files given as argument
-# @arg $@ files:String[] the files to filter
-# @exitcode * if one of the filter command fails
-# @stdin you can use stdin as alternative to files argument
-# @stdout the filtered content
-# @see https://en.wikipedia.org/wiki/ANSI_escape_code
-# shellcheck disable=SC2120
-Filters::removeAnsiCodes() {
-  # cspell:disable
-  sed -E 's/\x1b\[[0-9;]*[mGKHF]//g' "$@"
-  # cspell:enable
-}
-
 # @description log message to file
 # @arg $1 message:String the message to display
 Log::logDebug() {