From 4f7270df65d7c79222ab29b8628f251eee16053b Mon Sep 17 00:00:00 2001
From: Laurent Bercot <ska-skaware@skarnet.org>
Date: Tue, 26 May 2020 14:25:08 +0000
Subject:  New envfile format; prepare for 2.6.1.0

---
 NEWS                     |  9 +++++
 doc/envfile.html         | 40 +++++++++++++++------
 doc/index.html           |  2 +-
 doc/upgrade.html         | 10 ++++++
 package/info             |  2 +-
 src/execline/envfile.c   | 94 ++++++++++++++++++++++++++++++++++++++----------
 src/execline/envfile.txt | 60 +++++++++++++++++++++++++++++++
 7 files changed, 186 insertions(+), 31 deletions(-)
 create mode 100644 src/execline/envfile.txt

diff --git a/NEWS b/NEWS
index d966ede..07f4fef 100644
--- a/NEWS
+++ b/NEWS
@@ -1,5 +1,14 @@
 Changelog for execline.
 
+In 2.6.1.0
+----------
+
+ - The envfile format has changed; it is now more expressive and
+largely compatible with systemd's EnvironmentFile format. The
+new format is a superset of the old one, so old envfiles should
+still be understood by the new binary.
+
+
 In 2.6.0.2
 ----------
 
diff --git a/doc/envfile.html b/doc/envfile.html
index 5c0523c..713b6ae 100644
--- a/doc/envfile.html
+++ b/doc/envfile.html
@@ -55,19 +55,32 @@ it execs into <em>prog</em>.
 
 <p>
  <em>file</em> is a text file containing lines of the form <tt>key = value</tt>.
-Whitespace is permitted before and after <em>key</em>, and before or after <em>value</em>,
-but <em>key</em> or <em>value</em> can never contain whitespace. No quoting
-is possible.
+Whitespace is permitted before and after <em>key</em>, and before or after <em>value</em>.
 </p>
 
 <p>
  Empty lines, or lines containing only whitespace, are ignored.
 Lines beginning with <tt>#</tt> (possibly after some whitespace)
 are ignored (and typically used for comments).
-Comments are also possible at the end of lines:
-<tt>key = value # comment</tt> is a valid line. Note that there
-<em>must</em> be whitespace between <em>value</em> and <tt>#</tt>
-in this case (else <tt>#</tt> is just read as a part of <em>value</em>).
+Leading and trailing whitespace is stripped from values; but a
+value can be double-quoted, which allows for inclusion of leading
+and trailing whitespace.
+</p>
+
+<p>
+ A non-commented line that ends with a backslash ("<tt>\</tt>") is
+concatenated with the following one, and the newline character is
+ignored. If a backslashed newline happens before the start of a
+value, then the whitespace at the beginning of the new line will be
+part of the value (i.e. leading whitespace on a new line is not
+stripped).
+</p>
+
+<p>
+ C escapes, including hexadecimal and octal sequences, are supported
+in quoted values. Unicode codepoint sequences are not supported. It
+is possible to include a newline in a quoted value by using <tt>\n</tt>;
+but including newlines in environment variables is not recommended.
 </p>
 
 <p>
@@ -78,16 +91,21 @@ added to the environment, with an empty value. If you do not want
 the environment.
 </p>
 
+<p>
+ The envfile format is largely compatible with systemd's
+<a href="https://www.freedesktop.org/software/systemd/man/systemd.exec.html#EnvironmentFile=">EnvironmentFile</a>
+format, which allows for the reuse of such files outside of systemd.
+</p>
+
 <h2> Notes </h2>
 
 <ul>
  <li> If <em>file</em> is <tt>-</tt> then <tt>envfile</tt> reads
 and parses its standard input instead. To read a file literally named
 <tt>-</tt>, you can use <tt>./-</tt> for instance. </li>
- <li> The variables and values that can be defined with <tt>envfile</tt>
-are purposefully restricted in order to keep the parsing extremely simple.
-If you need fancy values, for instance values that contain whitespace or
-a newline, you should use an envdir instead: see
+ <li> The variables that can be defined with <tt>envfile</tt>
+are purposefully restricted. If you need more expressive power for your
+variable names, or for your values, you should use an envdir instead: see
 <a href="//skarnet.org/software/s6/s6-envdir.html">s6-envdir</a>. </li>
 </ul>
 
diff --git a/doc/index.html b/doc/index.html
index 0b41cea..0437fd5 100644
--- a/doc/index.html
+++ b/doc/index.html
@@ -66,7 +66,7 @@ library. </li>
 <h3> Download </h3>
 
 <ul>
- <li> The current released version of execline is <a href="execline-2.6.0.2.tar.gz">2.6.0.2</a>. </li>
+ <li> The current released version of execline is <a href="execline-2.6.1.0.tar.gz">2.6.1.0</a>. </li>
  <li> Alternatively, you can checkout a copy of the
 <a href="//git.skarnet.org/cgi-bin/cgit.cgi/execline/">execline
 git repository</a>:
diff --git a/doc/upgrade.html b/doc/upgrade.html
index e1ae3a3..3038dcf 100644
--- a/doc/upgrade.html
+++ b/doc/upgrade.html
@@ -18,6 +18,16 @@
 
 <h1> What has changed in execline </h1>
 
+<h2> in 2.6.1.0 </h2>
+
+<ul>
+ <li> The <a href="envfile.html">envfile</a> format has changed.
+It is now more expressive and largely compatible with systemd's
+<a href="https://www.freedesktop.org/software/systemd/man/systemd.exec.html#EnvironmentFile=">EnvironmentFile</a>
+format. The new format is a superset of the old one, so old
+envfiles should still be understood by the new binary. </li>
+</ul>
+
 <h2> in 2.6.0.2 </h2>
 
 <ul>
diff --git a/package/info b/package/info
index 22fb0b7..dcded8e 100644
--- a/package/info
+++ b/package/info
@@ -1,4 +1,4 @@
 package=execline
-version=2.6.0.2
+version=2.6.1.0
 category=admin
 package_macro_name=EXECLINE
diff --git a/src/execline/envfile.c b/src/execline/envfile.c
index 638b182..18a175b 100644
--- a/src/execline/envfile.c
+++ b/src/execline/envfile.c
@@ -3,9 +3,11 @@
 #include <stdint.h>
 #include <string.h>
 
+#include <skalibs/bytestr.h>
 #include <skalibs/types.h>
 #include <skalibs/buffer.h>
 #include <skalibs/strerr2.h>
+#include <skalibs/fmtscan.h>
 #include <skalibs/env.h>
 #include <skalibs/stralloc.h>
 #include <skalibs/djbunix.h>
@@ -14,6 +16,15 @@
 #define dieusage() strerr_dieusage(100, USAGE)
 #define dienomem() strerr_diefu1sys(111, "stralloc_catb")
 
+static void scanoct (stralloc *sa, size_t pos)
+{
+  unsigned int u ;
+  if (!stralloc_0(sa)) dienomem() ;
+  uint_oscan(sa->s + pos, &u) ;
+  sa->s[pos] = u ;
+  sa->len = pos+1 ;
+}
+
 static inline uint8_t cclass (char c)
 {
   switch (c)
@@ -26,7 +37,38 @@ static inline uint8_t cclass (char c)
     case '\t' :
     case '\f' :
     case '\r' : return 4 ;
-    default : return 5 ;
+    case '\\' : return 5 ;
+    case '\"' : return 6 ;
+    case 'a' :
+    case 'b' :
+    case 'f' : return 7 ;
+    case 'n' :
+    case 'r' :
+    case 't' :
+    case 'v' : return 8 ;
+    case '\'' :
+    case '?' : return 9 ;
+    case 'x' : return 10 ;
+    case '0' :
+    case '1' :
+    case '2' :
+    case '3' :
+    case '4' :
+    case '5' :
+    case '6' :
+    case '7' : return 11 ;
+    case '8' :
+    case '9' :
+    case 'A' :
+    case 'B' :
+    case 'c' :
+    case 'C' :
+    case 'd' :
+    case 'D' :
+    case 'e' :
+    case 'E' :
+    case 'F' : return 12 ;
+    default : return 13 ;
   }
 }
 
@@ -39,32 +81,48 @@ static inline char next (char const *file, buffer *b)
   return c ;
 }
 
-static inline void parse_config (char const *file, buffer *b, stralloc *modif)
+static inline void parse_config (char const *file, buffer *b, stralloc *sa)
 {
-  static uint8_t const table[7][6] =
+  static uint16_t const table[14][14] =
   {
-    { 0x07, 0x01, 0x80, 0x08, 0x00, 0x22 },
-    { 0x07, 0x01, 0x80, 0x01, 0x01, 0x01 },
-    { 0x08, 0x08, 0x08, 0x24, 0x03, 0x22 },
-    { 0x08, 0x08, 0x08, 0x24, 0x03, 0x08 },
-    { 0x47, 0x41, 0xc0, 0x25, 0x04, 0x25 },
-    { 0x47, 0x41, 0xc0, 0x25, 0x46, 0x25 },
-    { 0x07, 0x01, 0x80, 0x08, 0x06, 0x08 }
+    { 0x000e, 0x0001, 0x0020, 0x000f, 0x0000, 0x000f, 0x000f, 0x0012, 0x0012, 0x000f, 0x0012, 0x0012, 0x0012, 0x0012 },
+    { 0x000e, 0x0001, 0x0020, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001, 0x0001 },
+    { 0x000f, 0x0012, 0x000f, 0x0014, 0x0003, 0x000f, 0x000f, 0x0012, 0x0012, 0x000f, 0x0012, 0x0012, 0x0012, 0x0012 },
+    { 0x000f, 0x000f, 0x000f, 0x0014, 0x0003, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f },
+    { 0x004e, 0x0015, 0x0060, 0x0015, 0x0004, 0x0008, 0x0007, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015 },
+    { 0x004e, 0x0015, 0x0060, 0x0015, 0x0116, 0x0008, 0x0007, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015 },
+    { 0x00ce, 0x0015, 0x00d0, 0x0015, 0x0016, 0x0008, 0x0007, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015, 0x0015 },
+    { 0x000f, 0x0017, 0x000f, 0x0017, 0x0017, 0x0009, 0x0005, 0x0017, 0x0017, 0x0017, 0x0017, 0x0017, 0x0017, 0x0017 },
+    { 0x000f, 0x000f, 0x0025, 0x000f, 0x0015, 0x0015, 0x0015, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f },
+    { 0x000f, 0x0017, 0x0027, 0x0017, 0x0017, 0x0017, 0x0017, 0x1017, 0x1017, 0x0017, 0x000a, 0x011c, 0x0017, 0x0017 },
+    { 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x001b, 0x000f, 0x000f, 0x000f, 0x001b, 0x001b, 0x000f },
+    { 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x000f, 0x0217, 0x000f, 0x000f, 0x000f, 0x0217, 0x0217, 0x000f },
+    { 0x000f, 0x0417, 0x000f, 0x0417, 0x0417, 0x0409, 0x0405, 0x0417, 0x0417, 0x0417, 0x0417, 0x001d, 0x0417, 0x0417 },
+    { 0x000f, 0x0417, 0x000f, 0x0417, 0x0417, 0x0409, 0x0405, 0x0417, 0x0417, 0x0417, 0x0417, 0x0817, 0x0417, 0x0417 }
   } ;
   unsigned int line = 1 ;
+  size_t mark = 0 ;
   uint8_t state = 0 ;
-  while (state < 7)
+  while (state < 14)
   {
     char c = next(file, b) ;
-    uint8_t what = table[state][cclass(c)] ;
+    uint16_t what = table[state][cclass(c)] ;
     state = what & 0x0f ;
-    if (what & 0x20)
-      if (!stralloc_catb(modif, &c, 1)) dienomem() ;
-    if (what & 0x40)
-      if (!stralloc_0(modif)) dienomem() ;
-    if (what & 0x80) line++ ;
+    if (what & 0x0400) scanoct(sa, mark) ;
+    if (what & 0x0100) mark = sa->len ;
+    if (what & 0x1000) c = 7 + byte_chr("abtnvfr", 7, c) ;
+    if (what & 0x0010) if (!stralloc_catb(sa, &c, 1)) dienomem() ;
+    if (what & 0x0020) line++ ;
+    if (what & 0x0080) sa->len = mark ;
+    if (what & 0x0040) if (!stralloc_0(sa)) dienomem() ;
+    if (what & 0x0200)
+    {
+      sa->s[sa->len-2] = (fmtscan_num(sa->s[sa->len-2], 16) << 4) + fmtscan_num(sa->s[sa->len-1], 16) ;
+      sa->len-- ;
+    }
+    if (what & 0x0800) scanoct(sa, mark) ;
   }
-  if (state > 7)
+  if (state > 14)
   {
     char fmt[UINT_FMT] ;
     fmt[uint_fmt(fmt, line)] = 0 ;
diff --git a/src/execline/envfile.txt b/src/execline/envfile.txt
new file mode 100644
index 0000000..81ed02e
--- /dev/null
+++ b/src/execline/envfile.txt
@@ -0,0 +1,60 @@
+envfile parser:
+
+class	|	0	1	2	3	4	5	6	7	8	9	10	11	12	13
+st\ev	|	\0	#	\n	=	space	\	"	abf	nrtv	'?	x	0-7	89cde	other
+
+START	|			n					a	a		a	a	a	a
+0	|	END	COMMENT	START	X	START	X	X	VAR	VAR	X	VAR	VAR	VAR	VAR
+
+COMMENT |			n
+1	|	END	COMMENT	START	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT	COMMENT
+
+VAR	|		a		a				a	a		a	a	a	a
+2	|	X	VAR	X	EQ	WAITEQ	X	X	VAR	VAR	X	VAR	VAR	VAR	VAR
+
+WAITEQ	|				a
+3	|	X	X	X	EQ	WAITEQ	X	X	X	X	X	X	X	X	X
+
+EQ	|	+	a	+ n	a				a	a	a	a	a	a	a
+4	|	END	VAL	START	VAL	EQ	ESC	Q	VAL	VAL	VAL	VAL	VAL	VAL	VAL
+
+VAL	|	+	a	+ n	a	m a			a	a	a	a	a	a	a
+5	|	END	VAL	START	VAL	SPACE	ESC	Q	VAL	VAL	VAL	VAL	VAL	VAL	VAL
+
+SPACE	|	b +	a	b + n	a	a			a	a	a	a	a	a	a
+6	|	END	VAL	START	VAL	SPACE	ESC	Q	VAL	VAL	VAL	VAL	VAL	VAL	VAL
+
+Q	|		a		a	a			a	a	a	a	a	a	a
+7	|	X	Q	X	Q	Q	QESC	VAL	Q	Q	Q	Q	Q	Q	Q
+
+ESC	|			n		a	a	a
+8	|	X	X	VAL	X	VAL	VAL	VAL	X	X	X	X	X	X	X
+
+QESC	|		a	n	a	a	a	a	s a	s a	a		m a	a	a
+9	|	X	Q	Q	Q	Q	Q	Q	Q	Q	Q	HEX0	OCT1	Q	Q
+
+HEX0	|								a				a	a
+a	|	X	X	X	X	X	X	X	HEX1	X	X	X	HEX1	HEX1	X
+
+HEX1	|								a h				a h	a h
+b	|	X	X	X	X	X	X	X	Q	X	X	X	Q	Q	X
+
+OCT1	|		o a		o a	o a	o	o	o a	o a	o a	o a	a	o a	o a
+c	|	X	Q	X	Q	Q	QESC	VAL	Q	Q	Q	Q	OCT2	Q	Q
+
+OCT2	|		o a		o a	o a	o	o	o a	o a	o a	o a	a O	o a	o a
+d	|	X	Q	X	Q	Q	QESC	VAL	Q	Q	Q	Q	Q	Q	Q
+
+END = e, X = f
+
+16 states, 9 actions -> 13 bits
+
+a	0x0010	add cur
+n	0x0020	line++
++	0x0040	add \0 (definition complete)
+b	0x0080	go back to mark
+m	0x0100	set mark
+h	0x0200	calc hex
+o	0x0400	calc oct before a
+O	0x0800	calc oct after a
+s	0x1000	turn cur special before a
-- 
cgit v1.3.1