配方似乎忽略條件內環

Question

我有我的make文件這個食譜：

.PHONY: matlab           
matlab: $(matlab_files)         
    set -x;for f in `echo "$^"`; do\      
    base=`basename $$f`;\        
    if [[ ! -f "$(matlab_dir_name)/$${base}" ]]; then\ 
     ln -s $$f $(matlab_dir_name)/$${base}; \   
    fi;\            
    done;set +x           

就是，我要檢查，如果鏈接使得它之前就存在。但是，如果測試似乎被忽略。通過上面顯示的調試選項，我得到以下幾行：

+ for f in '`echo "MatlabD3PDImport/matlab/d3pd_export.m MatlabD3PDImport/matlab/matd3pd.m MatlabD3PDImport/matlab/net2py.m MatlabRingerTrain/matlab/cluster_size.m MatlabRingerTrain/matlab/create_cluster.m MatlabRingerTrain/matlab/getEffAtThres.m MatlabRingerTrain/matlab/getThresAndEff.m MatlabRingerTrain/matlab/scanparam.m MatlabRingerTrain/matlab/script_normal_train_offline.m MatlabRingerTrain/matlab/trainBinaryClassifier.m MatlabRingerTrain/matlab/trainRinger.m MatlabRingerTrain/matlab/xValidateBinaryClassifier.m"`' 
++ basename MatlabRingerTrain/matlab/scanparam.m 
+ base=scanparam.m 
+ [[ ! -f matlab/scanparam.m ]] 
+ ln -s MatlabRingerTrain/matlab/scanparam.m matlab/scanparam.m 
ln: matlab/scanparam.m: File exists 
+ for f in '`echo "MatlabD3PDImport/matlab/d3pd_export.m MatlabD3PDImport/matlab/matd3pd.m MatlabD3PDImport/matlab/net2py.m MatlabRingerTrain/matlab/cluster_size.m MatlabRingerTrain/matlab/create_cluster.m MatlabRingerTrain/matlab/getEffAtThres.m MatlabRingerTrain/matlab/getThresAndEff.m MatlabRingerTrain/matlab/scanparam.m MatlabRingerTrain/matlab/script_normal_train_offline.m MatlabRingerTrain/matlab/trainBinaryClassifier.m MatlabRingerTrain/matlab/trainRinger.m MatlabRingerTrain/matlab/xValidateBinaryClassifier.m"`' 
++ basename MatlabRingerTrain/matlab/script_normal_train_offline.m 
+ base=script_normal_train_offline.m 
+ [[ ! -f matlab/script_normal_train_offline.m ]] 
+ ln -s MatlabRingerTrain/matlab/script_normal_train_offline.m matlab/script_normal_train_offline.m 
ln: matlab/script_normal_train_offline.m: File exists 
+ for f in '`echo "MatlabD3PDImport/matlab/d3pd_export.m MatlabD3PDImport/matlab/matd3pd.m MatlabD3PDImport/matlab/net2py.m MatlabRingerTrain/matlab/cluster_size.m MatlabRingerTrain/matlab/create_cluster.m MatlabRingerTrain/matlab/getEffAtThres.m MatlabRingerTrain/matlab/getThresAndEff.m MatlabRingerTrain/matlab/scanparam.m MatlabRingerTrain/matlab/script_normal_train_offline.m MatlabRingerTrain/matlab/trainBinaryClassifier.m MatlabRingerTrain/matlab/trainRinger.m MatlabRingerTrain/matlab/xValidateBinaryClassifier.m"`' 
++ basename MatlabRingerTrain/matlab/trainBinaryClassifier.m 
+ base=trainBinaryClassifier.m 
+ [[ ! -f matlab/trainBinaryClassifier.m ]] 
+ ln -s MatlabRingerTrain/matlab/trainBinaryClassifier.m matlab/trainBinaryClassifier.m 
ln: matlab/trainBinaryClassifier.m: File exists 
+ for f in '`echo "MatlabD3PDImport/matlab/d3pd_export.m MatlabD3PDImport/matlab/matd3pd.m MatlabD3PDImport/matlab/net2py.m MatlabRingerTrain/matlab/cluster_size.m MatlabRingerTrain/matlab/create_cluster.m MatlabRingerTrain/matlab/getEffAtThres.m MatlabRingerTrain/matlab/getThresAndEff.m MatlabRingerTrain/matlab/scanparam.m MatlabRingerTrain/matlab/script_normal_train_offline.m MatlabRingerTrain/matlab/trainBinaryClassifier.m MatlabRingerTrain/matlab/trainRinger.m MatlabRingerTrain/matlab/xValidateBinaryClassifier.m"`' 
++ basename MatlabRingerTrain/matlab/trainRinger.m 
+ base=trainRinger.m 
+ [[ ! -f matlab/trainRinger.m ]] 
+ ln -s MatlabRingerTrain/matlab/trainRinger.m matlab/trainRinger.m 
ln: matlab/trainRinger.m: File exists 
+ for f in '`echo "MatlabD3PDImport/matlab/d3pd_export.m MatlabD3PDImport/matlab/matd3pd.m MatlabD3PDImport/matlab/net2py.m MatlabRingerTrain/matlab/cluster_size.m MatlabRingerTrain/matlab/create_cluster.m MatlabRingerTrain/matlab/getEffAtThres.m MatlabRingerTrain/matlab/getThresAndEff.m MatlabRingerTrain/matlab/scanparam.m MatlabRingerTrain/matlab/script_normal_train_offline.m MatlabRingerTrain/matlab/trainBinaryClassifier.m MatlabRingerTrain/matlab/trainRinger.m MatlabRingerTrain/matlab/xValidateBinaryClassifier.m"`' 
++ basename MatlabRingerTrain/matlab/xValidateBinaryClassifier.m 
+ base=xValidateBinaryClassifier.m 
+ [[ ! -f matlab/xValidateBinaryClassifier.m ]] 
+ ln -s MatlabRingerTrain/matlab/xValidateBinaryClassifier.m matlab/xValidateBinaryClassifier.m 
ln: matlab/xValidateBinaryClassifier.m: File exists 
+ set +x 

爲什麼shell忽略if測試？

Answer 1

-f FILE檢查文件是否存在並且是常規文件。要在測試中包含符號鏈接，請使用-h/-L（僅符號鏈接）或-e（所有文件）。

if [[ ! -e "$(matlab_dir_name)/$${base}" ]]; then