R从Google MyTracks中提取。kml文件

Question

我想使用R从一个数组中读取“when”值。由Google My tracks创建的kml文件（摘录如下）：

 ?xml version="1.0" encoding="UTF-8"?>
 <kml xmlns="http://www.opengis.net/kml/2.2"
 xmlns:gx="http://www.google.com/kml/ext/2.2"
 xmlns:atom="http://www.w3.org/2005/Atom">
 <Document>
 <open>1</open>
 <visibility>1</visibility>
 <name><![CDATA[2013-06-29 1:09pm]]></name>
 <atom:author><atom:name><![CDATA[Created by Google My Tracks on Android.]]></atom:name>     </atom:author>

 ...

 <gx:MultiTrack>
 <altitudeMode>absolute</altitudeMode>
 <gx:interpolate>1</gx:interpolate>
 <gx:Track>
 <when>2013-06-29T17:09:04.564Z</when>
 <gx:coord>-79.305048 43.710639 72.9000015258789</gx:coord>
 <when>2013-06-29T17:09:06.135Z</when>
 <gx:coord>-79.304971 43.710653 67.4000015258789</gx:coord>
 <when>2013-06-29T17:09:08.135Z</when>
 <gx:coord>-79.305193 43.710535 78.19999694824219</gx:coord>
 <when>2013-06-29T17:09:09.135Z</when>

节点“when”是“与位置相对应的时间值（在gx:coord元素中指定）”。“gx:coord”是“由经度、纬度和高度三个值组成的坐标值”。(https://developers.google.com/kml/documentation/kmlreference#gxtrack)

我想要的值的路径是：

 kml/Document/Placemark/gx:MultiTrack/gx:Track/when

发件人：xmlstarlet el“filename.kml”

我能够使用以下方法提取坐标和高度：

 coords <- xpathSApply(check, "//gx:coord", xmlValue)
 lat <- sapply(strsplit(as.character(coords)," "), "[",1)
 lon <- sapply(strsplit(as.character(coords)," "), "[",2)
 ele <- sapply(strsplit(as.character(coords)," "), "[",3)

但是我当时还没能得到。我想从文件中提取的是：

 17:09:04.564
 17:09:06.135
 17:09:08.135
 17:09:09.135

并将它们与坐标和标高对齐。

我试过：

timeStamp <- xpathSApply(check, "//gx:MultiTrack", xmlValue)

这就得到了一个可以解析的字符串，因为时间以“T”开始，以“Z”结束：

 [1] "absolute12013-06-29T17:09:04.564Z-79.305048 43.710639 72.90000152587892013-06-   29T17:09:06.135Z-79.304971 43.710653 67.40000152587892013-06-29T17:09:08.135Z-79.305193 43.710535 78.199996948242192013-06-29T17:09:09.135Z-79.305164 43.710592 77.699996948242192013-06-29T17:09:10.134Z-79.305097 43.710614 67.52013-06-29T17:09:11.137Z-79.305066 43.710572 

有什么好主意吗？提前感谢。

编辑-----

我不优雅的解决方案：

 file_name <- "2013-06-29 1-09pm.kml"
 library(XML)
 # read XML tree schema
 check <-xmlInternalTreeParse(file=file_name)
 library(gsubfn)
 # read kml file into a string 
 z <- xpathSApply(check, "//gx:MultiTrack", xmlValue)
 # find text bounded by (and including) T and Z
 x <- strapply(z,"T.+?Z")
 # unpack the resulting list
 x1 <- unlist(x)
 # get rid of the initial T
 x2 <- gsub("T", "", x1)
 # get rid of the trailing Z
 x3 <- gsub("Z", "", x2)
 # convert it to a time format
 time <- strptime(x3, "%H:%M:%OS")

Answer 1

有更好的html解析方法，但我看没有人发布过。这将起作用，但regexing通常不赞成html解析。

library(qdap)
x <- unlist(genXtract(dat, "<when>", "</when>"))
y <- unlist(genXtract(dat, "<gx:coord>", "</gx:coord>"))

x[sapply(x, function(x) !identical(x, character(0)))]
y[sapply(y, function(x) !identical(x, character(0)))]

## > x[sapply(x, function(x) !identical(x, character(0)))]
##       <when>  :  </when>14       <when>  :  </when>16 
## "2013-06-29T17:09:04.564Z" "2013-06-29T17:09:06.135Z" 
##       <when>  :  </when>18       <when>  :  </when>20 
## "2013-06-29T17:09:08.135Z" "2013-06-29T17:09:09.135Z" 
## > y[sapply(y, function(x) !identical(x, character(0)))]
##             <gx:coord>  :  </gx:coord>15 
##  "-79.305048 43.710639 72.9000015258789" 
##             <gx:coord>  :  </gx:coord>17 
##  "-79.304971 43.710653 67.4000015258789" 
##             <gx:coord>  :  </gx:coord>19 
## "-79.305193 43.710535 78.19999694824219"