From 5bcb7b4ea057614cd041db696f43935f6f6a93f1 Mon Sep 17 00:00:00 2001
From: Jonathan Lebon <jlebon@redhat.com>
Date: Thu, 31 Oct 2013 15:38:29 -0400
Subject: [PATCH] levensh suggest unresolved probe points

This patch adds Levenshtein suggestion to match_node::find_and_build().
Suggestions are made in three cases:
   - Unresolved double-glob (e.g. syscall.poen**)
   - Unresolved glob (e.g. syscall.poen*)
   - Unresolved no-glob (e.g. kernel.functoin("bla"))

Note that suggestions for globby functors are based on the prefix part
only (part of functor prior to the **) to increase meaningfulness.

Example outputs:

syscall.poen --> similar: open, chown, fork, iopl, mount
kernel.functoin --> similar: function, data, trace, mark, statement

syscall.poen** --> similar: open, chown, fork, iopl, mount
syscall.poen* --> similar: open, chown, fork, iopl, mount
syscall.*poen --> similar: open, chown, ppoll, capget, capset
syscall.**poen --> similar: open, capget, capset, chown, fchown

Non-globby functors in general give good results. Globby stuff can be
troublesome:

sys**poen --> similar: nfs, vfs, _nfs, _vfs, end
---
 elaborate.cxx | 53 +++++++++++++++++++++++++++++++--------------------
 elaborate.h   |  1 +
 2 files changed, 33 insertions(+), 21 deletions(-)

diff --git a/elaborate.cxx b/elaborate.cxx
index 00ce8c9a2..01d03c8f6 100644
--- a/elaborate.cxx
+++ b/elaborate.cxx
@@ -541,12 +541,10 @@ match_node::find_and_build (systemtap_session& s,
         {
           // We didn't find any wildcard matches (since the size of
           // the result vector didn't change).  Throw an error.
-          string alternatives;
-          for (sub_map_iterator_t i = sub.begin(); i != sub.end(); i++)
-            alternatives += string(" ") + i->first.str();
-
-          throw SEMANTIC_ERROR(_F("probe point mismatch (alternatives: %s)",
-                                  alternatives.c_str()), comp->tok);
+          string sugs = suggest_functors(functor);
+          throw SEMANTIC_ERROR (_F("probe point mismatch: didn't find any wildcard matches%s",
+                                   sugs.empty() ? "" : (" (similar: " + sugs + ")").c_str()),
+                                comp->tok);
         }
     }
   else if (isglob(loc->components[pos]->functor)) // wildcard?
@@ -618,13 +616,10 @@ match_node::find_and_build (systemtap_session& s,
         {
 	  // We didn't find any wildcard matches (since the size of
 	  // the result vector didn't change).  Throw an error.
-          string alternatives;
-          for (sub_map_iterator_t i = sub.begin(); i != sub.end(); i++)
-            alternatives += string(" ") + i->first.str();
-
-          throw SEMANTIC_ERROR(_F("probe point mismatch %s didn't find any wildcard matches",
-                                  (alternatives == "" ? "" : _(" (alternatives: ") +
-                                   alternatives + ")").c_str()), loc->components[pos]->tok);
+          string sugs = suggest_functors(loc->components[pos]->functor);
+          throw SEMANTIC_ERROR (_F("probe point mismatch: didn't find any wildcard matches%s",
+                                   sugs.empty() ? "" : (" (similar: " + sugs + ")").c_str()),
+                                loc->components[pos]->tok);
 	}
     }
   else
@@ -657,18 +652,34 @@ match_node::find_and_build (systemtap_session& s,
         {
           // We didn't find any alias suffixes (since the size of the
           // result vector didn't change).  Throw an error.
-          string alternatives;
-          for (sub_map_iterator_t i = sub.begin(); i != sub.end(); i++)
-            alternatives += string(" ") + i->first.str();
-
-          throw SEMANTIC_ERROR(_F("probe point mismatch %s",
-                                  (alternatives == "" ? "" : (_(" (alternatives:") + alternatives +
-                                  ")").c_str())),
-                               loc->components[pos]->tok);
+          string sugs = suggest_functors(loc->components[pos]->functor);
+          throw SEMANTIC_ERROR (_F("probe point mismatch%s",
+                                   sugs.empty() ? "" : (" (similar: " + sugs + ")").c_str()),
+                                loc->components[pos]->tok);
         }
     }
 }
 
+string
+match_node::suggest_functors(string functor)
+{
+  // only use prefix if globby (and prefix is non-empty)
+  size_t glob = functor.find('*');
+  if (glob != string::npos && glob != 0)
+    functor.erase(glob);
+  if (functor.empty())
+    return "";
+
+  set<string> functors;
+  for (sub_map_iterator_t i = sub.begin(); i != sub.end(); i++)
+    {
+      string ftor = i->first.str();
+      if (ftor.find('(') != string::npos)  // trim any parameter
+        ftor.erase(ftor.find('('));
+      functors.insert(ftor);
+    }
+  return levenshtein_suggest(functor, functors, 5); // print top 5
+}
 
 void
 match_node::try_suffix_expansion (systemtap_session& s,
diff --git a/elaborate.h b/elaborate.h
index e93723d60..08e3d1f86 100644
--- a/elaborate.h
+++ b/elaborate.h
@@ -330,6 +330,7 @@ match_node
   void find_and_build (systemtap_session& s,
                        probe* p, probe_point *loc, unsigned pos,
                        std::vector<derived_probe *>& results);
+  std::string suggest_functors(std::string functor);
   void try_suffix_expansion (systemtap_session& s,
                              probe *p, probe_point *loc, unsigned pos,
                              std::vector<derived_probe *>& results);
-- 
2.43.5