Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-11-25 22:01:30 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

add a little more info to the readme file

Browse Source
This commit is contained in:
jrgm%netscape.com 2003-03-29 01:50:12 +00:00
parent a954c467fc
commit f96eb6fb33
1 changed files with 63 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

78
tools/page-loader/README.txt
View File

@ -19,7 +19,7 @@
# Contributor(s):
# John Morrison <jrgm@netscape.com>, original author
#
#
Rough notes on setting up this test app. jrgm@netscape.com 2001/08/05
1) this is intended to be run as a mod_perl application under an Apache web
@ -28,9 +28,9 @@ Rough notes on setting up this test app. jrgm@netscape.com 2001/08/05
page load].
2) it should be possible to run this under Apache on win32, but I expect that
there are *nix-oriented assumptions that have crept in. You would also need
there are *nix-oriented assumptions that have crept in. (You would also need
a replacement for Time::HiRes, probably by using Win32::API to directly
call into the system to Windows 'GetLocalTime()'.
call into the system to Windows 'GetLocalTime()'.)
3) You need to have a few "non-standard" Perl Modules installed. This script
will tell you which ones are not installed (let me know if I have left some
@ -58,6 +58,41 @@ for (@modules) {
Download the .tar.gz files you need, and then (for the most part) just
do 'perl Makefile.PL; make; make test; make install'.
[Update: 28-Mar-2003] I recently installed Redhat 7.2, as server, which
installed Apache 1.3.20 with mod_perl 1.24 and perl 5.6.0. I then ran the
CPAN shell (`perl -MCPAN -e shell') and after completing configuration, I
did 'install Bundle::CPAN', 'install Bundle::LWP' and 'install DBI' to
upgrade tose modules and their dependencies.
CGI::Request seems to have disappeared from CPAN, but you can get a copy
from <http://stein.cshl.org/WWW/software/CGI::modules/> and then install
with the standard `perl Makefile.PL; make; make test; make install'.
To install the SQL::Statement, Text::CSV_XS, and DBD::CSV modules, there is
a bundle available on CPAN, so you can use the CPAN shell and just enter
'install Bundle::DBD::CSV'.
At the end of this, the output for the test program above was the
following. (Note: you don't necessarily have to have the exact version
numbers for these modules, as far as I know, but something close would be
safest).
LWP::UserAgent, version: 2.003
SQL::Statement, version: 1.005
Text::CSV_XS, version: 0.23
DBD::CSV, version: 0.2002
DBI, version: 1.35
Time::HiRes, version: 1.43
CGI::Request, version: 2.75
URI, version: 1.23
MIME::Base64, version: 2.18
HTML::Parser, version: 3.27
HTML::Tagset, version: 3.03
Digest::MD5, version: 2.24
I've also heard that installing with Apache 2.x requires a few changes,
although I don't know what they are. If you know, tell jrgm@netscape.com.
4) There is code to draw a sorted graph of the final results, but I have
disabled the place in 'report.pl' where its use would be triggered (look
for the comment). This is so that you can run this without having gone
@ -65,6 +100,15 @@ for (@modules) {
GD::Graph. If you have those in place, you can turn this on by just
reenabling the print statement in report.pl
[Note - 28-Mar-2003: with Redhat 7.2, libgd.so.1.8.4 is preinstalled to
/usr/lib. The current GD.pm modules require libgd 2.0.5 or higher, but you
use 1.8.4 if you install GD.pm version 1.40, which is available at
<http://stein.cshl.org/WWW/software/GD/old/GD-1.40.tar.gz>. Just do 'perl
Makefile.PL; make; make install' as usual. I chose to build with JPEG
support, but without FreeType, XPM and GIF support. I had a test error when
running 'make test', but it works fine for my purposes. I then installed
'GD::Text' and 'GD::Graph' from the CPAN shell.]
5) To set this up with Apache, create a directory in the cgi-bin for the web
server called e.g. 'page-loader' and then place this in the Apache
httpd.conf file to enable this for mod_perl (and then restart Apache).
@ -86,18 +130,22 @@ Options +ExecCGI
executes as (e.g., 'nobody' or 'apache'). [You may want to figure out some
other way to do this if this web server is not behind a firewall].
7) You need to assemble a set of content pages, with all images, included JS and
CSS pulled to the same directory. These pages can live anywhere on the same
HTTP server that is running this app. The app assumes that each page is in
its own sub-directory, with included content below that directory. You can
set the location and the list of pages in the file 'urllist.txt'. There are
various tools that will pull in complete copies of web pages (e.g. 'wget' or
something handrolled from LWP::UserAgent). You should edit the pages to remove
any redirects, popup windows, and possibly any platform specific JS rules
(e.g., Mac specific CSS included with 'document.write("LINK...'). You should
also check that for missing content, or URLs that did not get changed to point
to the local content. [One way to check for this is tweak this simple proxy server
to check your links: http://www.stonehenge.com/merlyn/WebTechniques/col34.listing.txt)
7) You need to assemble a set of content pages, with all images, included JS
and CSS pulled to the same directory. These pages can live anywhere on the
same HTTP server that is running this app. The app assumes that each page
is in its own sub-directory, with included content below that
directory. You can set the location and the list of pages in the file
'urllist.txt'. [See 'urllist.txt' for further details on what needs to be
set there.]
There are various tools that will pull in complete copies of web pages
(e.g. 'wget' or something handrolled from LWP::UserAgent). You should edit
the pages to remove any redirects, popup windows, and possibly any platform
specific JS rules (e.g., Mac specific CSS included with
'document.write("LINK...'). You should also check that for missing content,
or URLs that did not get changed to point to the local content. [One way to
check for this is tweak this simple proxy server to check your links:
http://www.stonehenge.com/merlyn/WebTechniques/col34.listing.txt)
8) The "hook" into the content is a single line in each top-level document like this:
<!-- MOZ_INSERT_CONTENT_HOOK -->