From f3e0a2a8008be1fbe08ace9b9da1090ee73bd79d Mon Sep 17 00:00:00 2001
From: Federico Ceratto <federico.ceratto@gmail.com>
Date: Tue, 24 Oct 2017 20:30:48 +0100
Subject: [PATCH] Add normalizePath, absolutePath and tests

---
 lib/pure/os.nim      | 63 ++++++++++++++++++++++++++++++++++++++++++--
 tests/stdlib/tos.nim | 50 +++++++++++++++++++++++++++++++++++
 2 files changed, 111 insertions(+), 2 deletions(-)

diff --git a/lib/pure/os.nim b/lib/pure/os.nim
index a77bee99daa95..487329c2e3733 100644
--- a/lib/pure/os.nim
+++ b/lib/pure/os.nim
@@ -288,8 +288,11 @@ proc setCurrentDir*(newDir: string) {.inline, tags: [].} =
 
 proc expandFilename*(filename: string): string {.rtl, extern: "nos$1",
   tags: [ReadDirEffect].} =
-  ## Returns the full (`absolute`:idx:) path of the file `filename`,
-  ## raises OSError in case of an error.
+  ## Returns the full (`absolute`:idx:) path of an existing file `filename`,
+  ## raises OSError in case of an error. Follows symlinks.
+  ##
+  ## To create absolute paths from any path (existing or not) see
+  ## `<#absolutePath>`_.
   when defined(windows):
     var bufsize = MAX_PATH.int32
     when useWinUnicode:
@@ -328,6 +331,62 @@ proc expandFilename*(filename: string): string {.rtl, extern: "nos$1",
       result = $r
       c_free(cast[pointer](r))
 
+proc normalizePath*(path: var string) {.rtl, extern: "nos$1", tags: [].} =
+  ## Normalize an UNIX path.
+  ##
+  ## Consecutive slashes (/) are collapsed, including an initial double slash.
+  ##
+  ## On relative paths, double dot (..) sequences are collapsed if possible.
+  ## On absolute paths they are always collapsed.
+  ##
+  ## Warning: URL-encoded and Unicode attempts at directory traversal are not detected.
+  ## Triple dot is not handled.
+  let is_abs = path[0] == DirSep
+  var stack: seq[string] = @[]
+  for p in split(path, {DirSep}):
+    case p
+    of "", ".":
+      continue
+    of "..":
+      if stack.len == 0:
+        if is_abs:
+          discard  # collapse all double dots on absoluta paths
+        else:
+          stack.add(p)
+      elif stack[^1] == "..":
+        stack.add(p)
+      else:
+        discard stack.pop()
+    else:
+      stack.add(p)
+
+  if is_abs:
+    path = DirSep & join(stack, $DirSep)
+  elif stack.len > 0:
+    path = join(stack, $DirSep)
+  else:
+    path = "."
+
+proc normalizedPath*(path: string): string {.rtl, extern: "nos$1", tags: [].} =
+  ## Returns a normalized UNIX path. See `<#normalizePath>`_
+  result = path
+  normalizePath result
+
+proc makeAbsolutePath*(path: var string) {.rtl, extern: "nos$1", tags: [].} =
+  ## Generates the normalized, (`absolute`:idx:) version of `path`, based on
+  ## the current directory. See `<#normalizePath>`_
+  if path.isAbsolute == false:
+    path = joinPath(getCurrentDir(), path)
+
+  if path.len > 1 and path[^1] == DirSep:
+    path = path[0..^2]
+
+proc absolutePath*(path: string): string {.rtl, extern: "nos$1", tags: [].} =
+  ## Returns the normalized, (`absolute`:idx:) version of `path`, based on
+  ## the current directory. See `<#normalizePath>`_
+  result = path
+  makeAbsolutePath result
+
 when defined(Windows):
   proc openHandle(path: string, followSymlink=true): Handle =
     var flags = FILE_FLAG_BACKUP_SEMANTICS or FILE_ATTRIBUTE_NORMAL
diff --git a/tests/stdlib/tos.nim b/tests/stdlib/tos.nim
index 771dc24562ba8..1ca6f822e86a8 100644
--- a/tests/stdlib/tos.nim
+++ b/tests/stdlib/tos.nim
@@ -42,6 +42,7 @@ Raises
 true
 true
 true
+
 '''
 """
 # test os path creation, iteration, and deletion
@@ -129,3 +130,52 @@ echo fileExists("../dest/a/b/file.txt")
 
 echo fileExists("../dest/a/b/c/fileC.txt")
 removeDir("../dest")
+
+block normalizedPath:
+  block relative:
+    doAssert normalizedPath(".") == "."
+    doAssert normalizedPath("..") == ".."
+    doAssert normalizedPath("../") == ".."
+    doAssert normalizedPath("../..") == "../.."
+    doAssert normalizedPath("../a/..") == ".."
+    doAssert normalizedPath("../a/../") == ".."
+    doAssert normalizedPath("./") == "."
+
+  block absolute:
+    doAssert normalizedPath("/") == "/"
+    doAssert normalizedPath("/.") == "/"
+    doAssert normalizedPath("/..") == "/"
+    doAssert normalizedPath("/../") == "/"
+    doAssert normalizedPath("/../..") == "/"
+    doAssert normalizedPath("/../../") == "/"
+    doAssert normalizedPath("/../../../") == "/"
+    doAssert normalizedPath("/./") == "/"
+    doAssert normalizedPath("//") == "/"
+    doAssert normalizedPath("///") == "/"
+    doAssert normalizedPath("/a//b") == "/a/b"
+    doAssert normalizedPath("/a///b") == "/a/b"
+    doAssert normalizedPath("/a/b/c/..") == "/a/b"
+    doAssert normalizedPath("/a/b/c/../") == "/a/b"
+
+when defined(Linux) or defined(osx):
+
+  block absolutePath:
+    block root:
+      let cwd = getCurrentDir()
+      setCurrentDir("/")
+      doAssert absolutePath("foo") == "/foo"
+      doAssert absolutePath("") == "/"
+      #doAssert absolutePath(".") == "/"
+      doAssert absolutePath("/") == "/"
+      setCurrentDir(cwd)
+
+    block tmp:
+      let cwd = getCurrentDir()
+      setCurrentDir("/tmp")
+      doAssert absolutePath("foo") == "/tmp/foo"
+      doAssert absolutePath("") == "/tmp"
+      #doAssert absolutePath(".") == "/tmp"
+      doAssert absolutePath("/") == "/"
+      #doAssert absolutePath("foo/..") == "/tmp/foo"
+      #doAssert absolutePath("../..") == "/tmp/foo"
+      setCurrentDir(cwd)