Python 使用正则表达式进行混乱文件名的模式匹配

我对REs没有太多经验，但需要解析100个文件名来生成“元数据”数据集。我已经能够生成包含文件路径和文件名的文本文件。解析出完整的文件名对我来说很简单，但我需要能够从文件名中解析出“sampleid”

问题是“样本ID”的语法到处都是（参见随附的csv示例数据：目标是从“样本”列转到“ID”列）。我尝试了一系列strsplit（）命令，但这非常麻烦，本质上不起作用。我还尝试过用一些基于语法结构的IF语句编写函数。我觉得这仍然不是一个好的解决方案，因为它仍然依赖于我在手之前手动识别不同的语法，而且我很容易错过一些东西，因为我必须用眼睛来做

在我看来，这是一个正则表达式问题，但我可以使用一些资源来帮助我开始。如果可能的话，我希望能够在R或Python中实现这一点。感谢您提供可能有用的任何资源或包/模块

dput(head(brain_ref, 25))
structure(list(file = c("/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/BXH12_1_brain_total_RNA_cDNA_GTCCGC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/BXH12_2_brain_total_RNA_cDNA_CAGATC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB13_1_brain_total_RNA_cDNA_ATGTCA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB13_2_brain_total_RNA_cDNA_GTGAAA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB17_1_brain_total_RNA_cDNA_CCGTCC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB17_2_brain_total_RNA_cDNA_ATGTCA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB2_1_brain_total_RNA_cDNA_GTCCGC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB2_2_brain_total_RNA_cDNA_CTTGTA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB25_1_brain_total_RNA_cDNA_AGTTCC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB25_2_brain_total_RNA_cDNA_AGTCAA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB27_1_brain_total_RNA_cDNA_CGATGT.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB27_2_brain_total_RNA_cDNA_AGTTCC.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB7_1_brain_total_RNA_cDNA_ACAGTG.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/HXB7_2_brain_total_RNA_cDNA_AGTCAA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/SHR_1_brain_total_RNA_cDNA_GCCAAT.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch1/ensembl_v96/SHR_2_brain_total_RNA_cDNA_TGACCA.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/ACI-SegHsd-2-brain-total-RNA_S17.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/BXH2-3-brain-total-RNA_S4.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/BXH5-3-brain-total-RNA_S3.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/BXH8-3-brain-total-RNA_S5.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/Cop-CrCrl-2-brain-total-RNA_S10.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/Dark-Agouti-1-brain-total-RNA_S16.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/Dark-Agouti-2-brain-total-RNA_S13.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/F344-NCI-1-brain-total-RNA_S18.genes.results", 
"/data/rn6/quantitation/brainTotalRNA/RI/batch10/ensembl_v96/F344-NCI-2-brain-total-RNA_S15.genes.results"
), sample = c("BXH12_1_brain_total_RNA_cDNA_GTCCGC", "BXH12_2_brain_total_RNA_cDNA_CAGATC", 
"HXB13_1_brain_total_RNA_cDNA_ATGTCA", "HXB13_2_brain_total_RNA_cDNA_GTGAAA", 
"HXB17_1_brain_total_RNA_cDNA_CCGTCC", "HXB17_2_brain_total_RNA_cDNA_ATGTCA", 
"HXB2_1_brain_total_RNA_cDNA_GTCCGC", "HXB2_2_brain_total_RNA_cDNA_CTTGTA", 
"HXB25_1_brain_total_RNA_cDNA_AGTTCC", "HXB25_2_brain_total_RNA_cDNA_AGTCAA", 
"HXB27_1_brain_total_RNA_cDNA_CGATGT", "HXB27_2_brain_total_RNA_cDNA_AGTTCC", 
"HXB7_1_brain_total_RNA_cDNA_ACAGTG", "HXB7_2_brain_total_RNA_cDNA_AGTCAA", 
"SHR_1_brain_total_RNA_cDNA_GCCAAT", "SHR_2_brain_total_RNA_cDNA_TGACCA", 
"ACI-SegHsd-2-brain-total-RNA_S17", "BXH2-3-brain-total-RNA_S4", 
"BXH5-3-brain-total-RNA_S3", "BXH8-3-brain-total-RNA_S5", "Cop-CrCrl-2-brain-total-RNA_S10", 
"Dark-Agouti-1-brain-total-RNA_S16", "Dark-Agouti-2-brain-total-RNA_S13", 
"F344-NCI-1-brain-total-RNA_S18", "F344-NCI-2-brain-total-RNA_S15"
), batch = c("batch1", "batch1", "batch1", "batch1", "batch1", 
"batch1", "batch1", "batch1", "batch1", "batch1", "batch1", "batch1", 
"batch1", "batch1", "batch1", "batch1", "batch10", "batch10", 
"batch10", "batch10", "batch10", "batch10", "batch10", "batch10", 
"batch10"), ID = c("BXH12_1", "BXH12_2", "HXB13_1", "HXB13_2", 
"HXB17_1", "HXB17_2", "HXB2_1", "HXB2_2", "HXB25_1", "HXB25_2", 
"HXB27_1", "HXB27_2", "HXB7_1", "HXB7_2", "SHR_1", "SHR_2", "ACI-SegHsd_2", 
"BXH2_3", "BXH5_3", "BXH8_3", "Cop-CrCrl_2", "Dark-Agouti_1", 
"Dark-Agouti_2", "F344-NCI_1", "F344-NCI_2")), row.names = c(NA, 
25L), class = "data.frame")

---

如果所有样本都包含

\u brain

或

-brain

，并且您希望在保存之前保存这些内容，您可以执行以下操作：

names=c（“ABBA-1_2-brain-total2”、“BABBA-2_2-brain-total2”、“ARA_1-1_2-brain-total2”）
gsub（“（.*.brain.*”，“\\1”，名称）
#>[1]“ABBA-1_2”“BABBA-2_2”“ARA_1-1_2”

灵感来自@r2evans:

---

samples在
R
中，我们还可以删除子字符串而不是捕获

sub("[-_]brain.*", "", names)
#[1] "ABBA-1_2"  "BABBA-2_2" "ARA_1-1_2"


或使用
trimws

trimws(names, whitespace = "[-_]brain.*")
#[1] "ABBA-1_2"  "BABBA-2_2" "ARA_1-1_2"


以OP公司为例

library(dplyr)
library(stringr)
brain_ref <-  brain_ref %>%
                 mutate(newsample = str_remove(sample,  "[_-]brain.*"))

库（dplyr）
图书馆（stringr）
大脑参考%
突变（newsample=str_remove（样本，[[u-]brain.*））

数据
名称为什么不：

for name in names:
    m = re.match(r'(.*)[_\-]brain[_\-]total', name)
    print(m.group(1))

你能在
R
中用
dput
而不是image
sub（“-（[0-9]+）-brain.*”、“\u1”，“SHR-1-brain-total-RNA\u S6”）
显示示例数据吗？@r2evans有时是下划线，而不是破折号。好的，
sub（[0-9]+）-brain.*”、“\u1”，“SHR-1-brain-total-RNA\u S6”）
（1）请不要发布代码/数据/错误的图像：它无法复制或搜索（SEO），它会破坏屏幕阅读器，并且可能不适合某些移动设备。参考：（和）。请直接包括代码或数据（例如，
dput（头（x））
或
data.frame（…）
）。（2） 即使你说有链接（而且没有），我还是建议链接过时，导致一个无法产生的问题。请按照akrun的建议，使用
dput
或编程方式（
data.frame
或
vec Try
“SHR-1-brain-total-RNA-S6”
）包含样本数据，应该返回
SHR_1
，但不是。（请参阅几分钟前我的评论。）我不确定你的意思。我尝试了
“SHR-1-brain-total-RNA-S6”
它确实返回了
SHR-1
，这是它应该返回的。我看不出有理由否决投票？对不起，我想这很清楚。
gsub（（.*）.brain.*，“\\1”，“SHR-1-brain-total-RNA-S6”）
返回
“SHR-1”
，但OP显示了
“SHR\u 1”
。所以当我说“应该返回”时，我是建议
“SHR-1”
是错误的。由于OP编辑了问题并更改了样本和期望值，因此不再有以
SHR-1
开头并期望
SHR_1
的字符串，因此我愿意取消对该问题的否决票。由于so不允许我撤回否决票，除非对答案进行编辑，因此我将您的答案编辑为（非常简单）更改一点，这样我就可以收回否决票。如果你愿意，请随意回滚我的编辑。谢谢。这几乎可以工作，但在数字靠近“大脑”的情况下（例如“F344”Stm_1braintotalRNAb30_S8_L001_R1_001”）@HarrySmith查看是否现在在第二个
[\u-]后面加上星号
和之前的
大脑
工作。
[\u-]*
现在表示模式的零次或多次重复。
names <- c("ABBA-1_2-brain-total2", "BABBA-2_2_brain-total2", "ARA_1-1_2-brain-total2")

for name in names:
    m = re.match(r'(.*)[_\-]brain[_\-]total', name)
    print(m.group(1))