Merge pull request #4145 from masatake/powershell-herestr

PowerShell: recognize herestrings

Units/parser-powershell.r/herestring.d/args.ctags

@@ -0,0 +1 @@
+--sort=no

Units/parser-powershell.r/herestring.d/expected.tags

@@ -0,0 +1,11 @@
+a	input.ps1	/^function a$/;"	f
+b	input.ps1	/^function b$/;"	f
+c	input-0.ps1	/^function c$/;"	f
+d	input-0.ps1	/^function d$/;"	f
+e	input-1.ps1	/^function e$/;"	f
+f	input-2.ps1	/^function f$/;"	f
+g	input-3.ps1	/^function g$/;"	f
+h	input-4.ps1	/^function h$/;"	f
+i	input-5.ps1	/^function i$/;"	f
+j	input-6.ps1	/^function j$/;"	f
+k	input-6.ps1	/^function k$/;"	f

Units/parser-powershell.r/herestring.d/input-0.ps1

@@ -0,0 +1,16 @@
+function c
+{
+    $i = 1
+    $heredoc0 =
+@'
+'"' $i
+'@
+    Write-Host $heredoc
+}
+function d
+{
+    Write-Host d
+}
+c
+d
+pause

Units/parser-powershell.r/herestring.d/input-1.ps1

@@ -0,0 +1,18 @@
+function e
+{
+    $i = 1
+    $heredoc0 =
+@'
+    Write-Host $heredoc
+}
+function inherestr
+{
+    Write-Host d
+}
+c
+d
+pause
+
+'"' $i
+'@
+}

Units/parser-powershell.r/herestring.d/input-2.ps1

@@ -0,0 +1,8 @@
+function f
+{
+    $i = 1
+    $heredoc0 =
+@'
+    Write-Host $heredoc
+@
+@'

Units/parser-powershell.r/herestring.d/input-3.ps1

@@ -0,0 +1,9 @@
+function g
+{
+    $i = 1
+    $heredoc0 =
+@'
+    Write-Host $heredoc
+@
+@
+

Units/parser-powershell.r/herestring.d/input-4.ps1

@@ -0,0 +1,8 @@
+function h
+{
+    $i = 1
+    $heredoc0 =
+@"
+    Write-Host $heredoc
+@
+@"

Units/parser-powershell.r/herestring.d/input-5.ps1

@@ -0,0 +1,13 @@
+function i
+{
+    $i = 1
+    $heredoc0 =
+@'
+    Write-Host $heredoc
+@
+@"
+}
+
+function donttagme
+{
+}

Units/parser-powershell.r/herestring.d/input-6.ps1

@@ -0,0 +1,12 @@
+function j
+{
+    $i = 1
+    $heredoc0 =
+@brokeninput
+# So ctags can do as it wants but the current implementation
+# intents extracting the next `k`.
+}
+
+function k
+{
+}

Units/parser-powershell.r/herestring.d/input.ps1

@@ -0,0 +1,16 @@
+function a
+{
+    $i = 1
+    $heredoc =
+@"
+""" $i
+"@
+    Write-Host $heredoc
+}
+function b
+{
+    Write-Host b
+}
+a
+b
+pause

parsers/powershell.c

@@ -9,6 +9,10 @@
 *
 *   This module contains code for generating tags for Windows PowerShell scripts
 *   (https://en.wikipedia.org/wiki/PowerShell).
+*
+*   References:
+*   - https://learn.microsoft.com/en-us/powershell/scripting/developer/windows-powershell-reference
+*   - https://learn.microsoft.com/en-us/powershell/scripting/lang-spec/chapter-01
 */
 
 /*
@@ -102,6 +106,7 @@ typedef struct {
 	unsigned long	lineNumber;
 	MIOPos			filePosition;
 	int				parentKind; /* KIND_GHOST_INDEX if none */
+	bool            herestring;	/* Meaningful only when type is TOKEN_STRING. */
 } tokenInfo;
 
 
@@ -258,6 +263,43 @@ static void parseString (vString *const string, const int delimiter)
 	}
 }
 
+static bool parseHereString (vString *const string, const int delimiter)
+{
+	enum phs_state {
+		PHS_INIT,
+		PHS_NEWLINE0,
+		PHS_QUOTE,
+		PHS_AT,
+		PHS_NEWLINE1,
+		PHS_END = PHS_NEWLINE1
+	} s = PHS_INIT;
+
+	while (s != PHS_END)
+	{
+		int c = getcFromInputFile ();
+
+		if (c == EOF)
+			return false;
+
+		vStringPut (string, c);
+
+		if (s == PHS_INIT && c == '\n')
+			s = PHS_NEWLINE0;
+		else if (s == PHS_NEWLINE0 && c == delimiter)
+			s = PHS_QUOTE;
+		else if (s == PHS_QUOTE && c == '@')
+			s = PHS_AT;
+		else if (s == PHS_AT && c == '\n')
+			s = PHS_NEWLINE1;
+		else
+			s = PHS_INIT;
+	}
+
+	Assert (vStringLength (string) >= 4);
+	vStringTruncate (string, vStringLength (string) - 4);
+	return true;
+}
+
 /* parse a identifier that may contain scopes, such as function names and
  * variable names.
  *
@@ -356,8 +398,33 @@ static void readTokenFull (tokenInfo *const token, bool includingScope)
 			parseString (token->string, c);
 			token->lineNumber = getInputLineNumber ();
 			token->filePosition = getInputFilePosition ();
+			token->herestring = false;
 			break;
 
+		case '@':
+		{
+			int d = getcFromInputFile ();
+			if (d == '\'' || d == '"')
+			{
+				bool is_herestr;
+				token->type = TOKEN_STRING;
+				is_herestr = parseHereString (token->string, d);
+				token->lineNumber = getInputLineNumber ();
+				token->filePosition = getInputFilePosition ();
+				if (is_herestr)
+					token->herestring = true;
+				else
+					token->type = TOKEN_UNDEFINED;
+			}
+			else if (d == EOF)
+				token->type = TOKEN_UNDEFINED;
+			else
+			{
+				ungetcToInputFile (d);
+				token->type = TOKEN_UNDEFINED;
+			}
+			break;
+		}
 		case '<':
 		{
 			int d = getcFromInputFile ();