From: blundellc <blundellc@edb9625f-4e0d-4859-8d74-9fd3b1da38cb>
Date: Sat, 13 Sep 2008 13:20:35 +0000 (+0000)
Subject: license: split out some of the heuristics for reducing poorly formed license fields.
X-Git-Url: https://git.donarmstrong.com/?a=commitdiff_plain;h=6e9d4006449c27d6c73e7371d2cedf3dd973f700;p=cran2deb.git

license: split out some of the heuristics for reducing poorly formed license fields.

in particular, do not remove numbers in just one step. it seems common
for people to use:
License: X11 (http://blahblah/dfgdfg/)
which will be reduce to just 'x' if these steps are not split.

Also separate license hashing into getting the license text and then
separately hashing via digest. (Useful for for the 'view' command about to
be added to cran2deb license.)

git-svn-id: svn://svn.r-forge.r-project.org/svnroot/cran2deb@76 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
---

diff --git a/pkg/trunk/R/license.R b/pkg/trunk/R/license.R
index c8a3dde..6453524 100644
--- a/pkg/trunk/R/license.R
+++ b/pkg/trunk/R/license.R
@@ -16,6 +16,12 @@ is_acceptable_license <- function(license) {
         message(paste('W: Accepting/rejecting wild license as',license,'. FIX THE PACKAGE!'))
         return(action)
     }
+    license <- license_text_extreme_reduce(license)
+    action = db_license_override_name(license)
+    if (!is.na(action)) {
+        message(paste('W: Accepting/rejecting wild license as',license,'. FIX THE PACKAGE!'))
+        return(action)
+    }
     # TODO: file {LICENSE,LICENCE} (+ maybe COPYING?)
     message(paste('E: Wild license',license,'did not match classic rules; rejecting'))
     return(F)
@@ -56,7 +62,13 @@ license_text_further_reduce <- function(license) {
     license = gsub('licen[sc]e','',license)
     license = gsub('(gnu )?(gpl|general public)','gpl',license)
     license = gsub('(mozilla )?(mpl|mozilla public)','mpl',license)
-    # remove everything that looks like a version specification
+    # remove any extra space introduced
+    license = chomp(gsub('[[:space:]]+',' ',license))
+    return(license)
+}
+
+license_text_extreme_reduce <- function(license) {
+    # remove everything that may or may not be a version specification
     license = gsub('(ver?sion|v)? *[0-9.-]+ *(or *(higher|later|newer|greater|above))?',''
                    ,license)
     # remove any extra space introduced
@@ -69,8 +81,8 @@ license_text_hash_reduce <- function(text) {
     return(chomp(tolower(gsub('[[:space:]]+',' ',text))))
 }
 
-get_license_hash <- function(pkg,license) {
-    license <- license_text_reduce(license)
+get_license <- function(pkg,license) {
+    license <- chomp(gsub('[[:space:]]+',' ',license))
     if (length(grep('^file ',license))) {
         if (length(grep('^file LICEN[CS]E$',license))) {
             path = gsub('file ','',license)
@@ -81,7 +93,11 @@ get_license_hash <- function(pkg,license) {
             return(NA)
         }
     }
-    return(digest(license,algo='sha1',serialize=FALSE))
+    return(license)
+}
+
+get_license_hash <- function(pkg,license) {
+    return(digest(get_license(pkg,license),algo='sha1',serialize=FALSE))
 }
 
 is_acceptable_hash_license <- function(pkg,license) {