From 2f04e7102eaad3159073019af96e6e5c4f2c9bcf Mon Sep 17 00:00:00 2001
From: Eelco Dolstra <e.dolstra@tudelft.nl>
Date: Mon, 16 Jun 2003 15:59:23 +0000
Subject: * Path hashing.

---
 src/hash.cc | 141 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 140 insertions(+), 1 deletion(-)

(limited to 'src/hash.cc')

diff --git a/src/hash.cc b/src/hash.cc
index 9451ac8d80e4..9558d36705f3 100644
--- a/src/hash.cc
+++ b/src/hash.cc
@@ -1,9 +1,16 @@
+#include <iostream>
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <dirent.h>
+#include <fcntl.h>
+
 extern "C" {
 #include "md5.h"
 }
 
 #include "hash.hh"
-#include <iostream>
 
 
 Hash::Hash()
@@ -88,3 +95,135 @@ Hash hashFile(const string & fileName)
     if (err) throw SysError("cannot hash file " + fileName);
     return hash;
 }
+
+
+struct HashSink : DumpSink
+{
+    struct md5_ctx ctx;
+    virtual void operator ()
+        (const unsigned char * data, unsigned int len)
+    {
+        md5_process_bytes(data, len, &ctx);
+    }
+};
+
+
+Hash hashPath(const string & path)
+{
+    Hash hash;
+    HashSink sink;
+    md5_init_ctx(&sink.ctx);
+    dumpPath(path, sink);
+    md5_finish_ctx(&sink.ctx, hash.hash);
+    return hash;
+}
+
+
+static void pad(unsigned int len, DumpSink & sink)
+{
+    if (len % 8) {
+        unsigned char zero[8];
+        memset(zero, 0, sizeof(zero));
+        sink(zero, 8 - (len % 8));
+    }
+}
+
+
+static void writeInt(unsigned int n, DumpSink & sink)
+{
+    unsigned char buf[8];
+    memset(buf, 0, sizeof(buf));
+    buf[0] = n & 0xff;
+    buf[1] = (n >> 8) & 0xff;
+    buf[2] = (n >> 16) & 0xff;
+    buf[3] = (n >> 24) & 0xff;
+    sink(buf, sizeof(buf));
+}
+
+
+static void writeString(const string & s, DumpSink & sink)
+{
+    unsigned int len = s.length();
+    writeInt(len, sink);
+    sink((const unsigned char *) s.c_str(), len);
+    pad(len, sink);
+}
+
+
+static void dumpEntries(const string & path, DumpSink & sink)
+{
+    DIR * dir = opendir(path.c_str());
+    if (!dir) throw SysError("opening directory " + path);
+    
+    struct dirent * dirent;
+
+    /* !!! sort entries */
+
+    while (errno = 0, dirent = readdir(dir)) {
+        string name = dirent->d_name;
+        if (name == "." || name == "..") continue;
+        writeString("entry", sink);
+        writeString("(", sink);
+        writeString("name", sink);
+        writeString(name, sink);
+        writeString("file", sink);
+        dumpPath(path + "/" + name, sink);
+        writeString(")", sink);
+    }
+
+    if (errno) throw SysError("reading directory " + path);
+    
+    closedir(dir); /* !!! close on exception */
+}
+
+
+static void dumpContents(const string & path, unsigned int size, 
+    DumpSink & sink)
+{
+    writeString("contents", sink);
+    writeInt(size, sink);
+
+    int fd = open(path.c_str(), O_RDONLY);
+    if (!fd) throw SysError("opening file " + path);
+    
+    unsigned char buf[16384];
+
+    unsigned int total = 0;
+    ssize_t n;
+    while ((n = read(fd, buf, sizeof(buf)))) {
+        if (n == -1) throw SysError("reading file " + path);
+        total += n;
+        sink(buf, n);
+    }
+
+    if (total != size)
+        throw SysError("file changed while reading it: " + path);
+
+    pad(size, sink);
+
+    close(fd); /* !!! close on exception */
+}
+
+
+void dumpPath(const string & path, DumpSink & sink)
+{
+    cerr << path << endl;
+
+    struct stat st;
+    if (lstat(path.c_str(), &st))
+        throw SysError("getting attributes of path " + path);
+
+    writeString("(", sink);
+
+    if (S_ISREG(st.st_mode)) {
+        writeString("type", sink);
+        writeString("regular", sink);
+        dumpContents(path, st.st_size, sink);
+    } else if (S_ISDIR(st.st_mode)) {
+        writeString("type", sink);
+        writeString("directory", sink);
+        dumpEntries(path, sink);
+    } else throw Error("unknown file type: " + path);
+
+    writeString(")", sink);
+}
-- 
cgit 1.4.1