From 464e0577d2a4b7e2584f487dbb4548e0f01e6152 Mon Sep 17 00:00:00 2001
From: Jacob Quinn <quinn.jacobd@gmail.com>
Date: Tue, 5 Feb 2019 08:31:05 -0700
Subject: [PATCH] Fix #154. We already have the ability to only read certain
 columns via manually iterating CSV.File or using Tables.select, but there was
 another issue where this file had an invalid column name that dies while
 trying to normalize the name.

---
 src/filedetection.jl                   | 2 +-
 test/testfiles/test_file_issue_154.csv | 3 +++
 test/testfiles/testfiles.jl            | 6 ++++++
 3 files changed, 10 insertions(+), 1 deletion(-)
 create mode 100644 test/testfiles/test_file_issue_154.csv

diff --git a/src/filedetection.jl b/src/filedetection.jl
index 990fc1b3..2ec6ac22 100644
--- a/src/filedetection.jl
+++ b/src/filedetection.jl
@@ -8,7 +8,7 @@ normalizename(name::Symbol) = name
 function normalizename(name::String)
     uname = strip(Unicode.normalize(name))
     id = Base.isidentifier(uname) ? uname : map(c->Base.is_id_char(c) ? c : '_', uname)
-    cleansed = string((!Base.is_id_start_char(id[1]) || id in RESERVED) ? "_" : "", id)
+    cleansed = string((isempty(id) || !Base.is_id_start_char(id[1]) || id in RESERVED) ? "_" : "", id)
     return Symbol(replace(cleansed, r"(_)\1+"=>"_"))
 end
 
diff --git a/test/testfiles/test_file_issue_154.csv b/test/testfiles/test_file_issue_154.csv
new file mode 100644
index 00000000..e12199b2
--- /dev/null
+++ b/test/testfiles/test_file_issue_154.csv
@@ -0,0 +1,3 @@
+a, b, ,
+0, 1, , comment 
+12, 5, ,
diff --git a/test/testfiles/testfiles.jl b/test/testfiles/testfiles.jl
index 440cdef0..b26f774b 100644
--- a/test/testfiles/testfiles.jl
+++ b/test/testfiles/testfiles.jl
@@ -490,4 +490,10 @@ testfiles = [
         NamedTuple{(:FAMILY, :PERSON, :MARKER, :RATIO), Tuple{String, String, String, Float64}},
         (FAMILY = ["A", "A", "A", "A", "A", "A", "EPGP013951", "EPGP014065", "EPGP014065", "EPGP014065", "EP07", "83346_EPGP014244", "83346_EPGP014244", "83506", "87001"], PERSON = ["EP01223", "EP01227", "EP01228", "EP01228", "EP01227", "EP01228", "EPGP013952", "EPGP014066", "EPGP014065", "EPGP014068", "706", "T3011", "T3231", "T17255", "301"], MARKER = ["rs710865", "rs11249215", "rs11249215", "rs10903129", "rs621559", "rs1514175", "rs773564", "rs2794520", "rs296547", "rs296547", "rs10927875", "rs2251760", "rs2251760", "rs2475335", "rs2413583"], RATIO = [0.0214, 0.0107, 0.00253, 0.0116, 0.00842, 0.0202, 0.00955, 0.0193, 0.0135, 0.0239, 0.0157, 0.0154, 0.0154, 0.00784, 0.0112])
     ),
+    # #154
+    ("test_file_issue_154.csv", (allowmissing=:auto, normalizenames=true),
+        (2, 4),
+        NamedTuple{(:a, :b, :_, :Column4), Tuple{Int64, Int64, Missing, Union{Missing, String}}},
+        (a = [0, 12], b = [1, 5], _ = Missing[missing, missing], Column4 = Union{Missing, String}[" comment ", missing])
+    ),
 ];